Full text of "A practical course in horology"

See other formats

PRACTICAL COURSE

HOROLOGY

From the collection of the

^ m

0 Prelinger
^ Jjibrary

D t P

San Francisco, California
2008

A PRACTICAL COURSE IN HOROLOGY

PRACTICAL COURSE

IN
HOROLOGY

By HAROLD C KELLY

Head, Department of Horology

Southwestern Institute of Technology

Weatherford, Oklahoma

THE MANUAL ARTS PRESS

PEORIA, ILLINOIS

THE MANUAL ARTS PRESS

No part of this book may be reproduced in any
form without permission of the copyright owner

34KS5

PRINTED IN THE UNITED STATES OF AMERICA

CONTENTS

"TIME'' Poem, by Laurens L, Simpson . . 7

Preface 9

PART I: GENERAL PRINCIPLES

Chapter

ONE. Wheel Work 13

TWO. Gearing 32

THREE. The Lever Escapement 39

FOUR. The Controlling Mechanism ... 52

PART II: PRACTICAL REPAIRING

ONE. Train Problems 62

TWO. Jeweling 6^

THREE. Making a Balance Staff 72>

FOUR. Pivoting 85

FIVE. Fitting Balance Springs 88

SIX. Escapement Adjusting 100

SEVEN. Cleaning and Oiling 124

PART III: ADJUSTING

ONE. Preliminary Notes on Adjusting . . 128

TWO. Position Adjusting 140

THREE. Adjustment to Isochronism . . . .150

FOUR. Adjustment to Temperature. . . .156

FIVE. Practical Work of Adjusting . . . 159

Glossary of Terms 173

Bibliography 182

Index 185

TIME

BY Laurens L. Simpson
(written to accompany the gift of a watch)

•

/ am time.

I come to give thee life
Twenty-four hours of every day,
All this to every man.

I tick tick.

I sound in peace, and strife,
In sleep, in work and play.
Forever, on and on.

I never change
To good and had alike.
The rich, the brave, the free
All use me as they may.

I am gold

To those who will; to others
Lead, who do not see
The benefit of industry.

I am power.

Weak to strong, coward to brave,

Man grows as I move on,

Or not, as will he may.

(continued next page)

I am fame

To those whose light is bright,
Who shine with all their might,
Pure, through day and night.

I am happiness
To those who serve and give,
Who help the weak, mxike known
The unknown, and live.

Now or never
Equal chance I give to all,
My days to use or lose.
Here once then gone forever.

PREFACE

The art of horology unquestionably ranks among the
most wonderful of the mechanical arts. One can only marvel
at the diminutive size of the modern wrist watch and the
accuracy of the machines by which the duplicate parts are
made.

Production and improved manufacturing methods have
also changed the repairman's approach to horology. Dup-
licate parts are available, so the horologist is seldom called
upon to make a part. However, since the sizes of watches
have been reduced, new tools and improved methods are
essential to good workmanship. One must develop a greater
skill in fitting staffs to small, uncut balance wheels, in adjust-
ing small escapements, and in handling the new, alloyed
balance springs.

The purpose of this book is to present the fundamentals
of horology, both in theory and practice. Part 1 deals with
wheel work and gearing, which involve the work of calculat-
ing the number of teeth of missing wheels and pinions and
in determining their proper diameters. Principles of escape-
ment design and an analysis of the balance and spring are
given considerable space. Part 2 treats repair methods, in
which the making of a balance staff and the adjustment of
the escapement are given more than the usual space allotted
to these subjects. Part 3 is concerned with the adjust-
ments to position, isochronism, and temperature, factors that

10 A PRACTICAL COURSE IN HOROLOGY

may be called the finishing touches of the horological pro-
fession.

The author is indebted to T. J. Wilkinson and C. E.
DeLong for the reading of parts of the manuscript and for
helpful suggestions. The chapter on wheel work is based
on a system by Jules Grossman, late director of the horolog-
ical school of Locle.

It is hoped that this work will contribute some small part
toward the development of a generation of capable and well
equipped horologists.

HAROLD C. KELLY

PART 1

GENERAL PRINCIPLES

Parti

CHAPTER ONE

Wheel Work

Terminology

Wheel work is the basis for the construction of all in-
struments for the registering of time. Even the clepsydra,
one of the earliest forms of a clock, which operated by the
dripping of water, involved the use of wheels. The wheels of
these early clocks as well as those of the first pendulum clocks
were filed out by hand and although the workmanship on some
of the later creations was quite skillfully executed, they were,
of course, very crude as compared with the machine-made
wheels of modern clocks and watches. Since wheels are
fundamental to the construction of watches, we have decided
that this opening chapter shall be given over to the considera-
tion of wheels as they apply to the science of horology. Be-
low are listed several definitions relative to the subject:

Wheel: any circular piece of metal on the periphery of
which teeth may be cut of various forms and numbers.

Pinion: the smaller wheel with teeth called leaves, work-
ing in connection with a larger wheel.

Train: a combination of two or more wheels and pinions,
geared together and transmitting power from one part of a
mechanism to another.

14 A PRACTICAL COURSE IN HOROLOGY

Barrel: a circular box of metal for the reception of the
main spring.

Balance: the vibratory wheel, which, in connection with
the balance spring, controls the movement of the hands.

Balance spring: a fine, coiled wire, one end of which is
attached by a collet to the balance staff and the other end to
some stationary part of the watch through the medium of a
piece called a stud.

Beat: one vibration of the balance and balance spring
resulting from an impulse received by means of an escape-
ment.

Escapement: includes those parts of a watch which
change the circular force of the escape wheel into the vibra-
tory motion of the balance.

Pallets: that part of an escapement which receives im-
pulse from the escape wheel and by means of a lever delivers
impulse to the balance. This term includes the pallet arms
and jewels.

Receiving pallet: that pallet stone over which a tooth
of the escape wheel slides in order to enter between the
pallet stones.

Discharging pallet: that pallet stone over which a tooth of
the escape wheel slides in order to leave from between the
pallets.

Lock: the overlapping contact of an escape wheel tooth
on a pallet stone's locking face.

Watch Trains
Since the motive force stored In the main spring of a
watch does not act directly on the balance, it is by necessity

WHEEL WORK 15

transmitted by a system of toothed wheels and pinions. This
system of wheels and pinions, commonly called a train, is a
scientific, mathematical assemblage of mobiles, and anyone
who becomes familiar with the principles involved derives
much satisfaction from such knowledge.

In watches we have two trains, the main train and the
dial train. The main train changes a slow motion into a
fast one with the particular purpose of causing the wheel
that carries the minute hand to make one turn in the same
time that the escapement makes a required number of beats.

The dial train, on the other hand, changes a fast motion
Into a slow one for the purpose of governing the distance the
hour hand travels to one turn of the minute hand.

The Main Train

Calculating the number of turns of a pinion. In order

to obtain the number of turns of a pinion into which a wheel

is geared, we divide the num- Q

ber of teeth in the wheel by the

number of leaves in the pinion.

Suppose, for example, a wheel

of 72 teeth gears into a pinion

of 12 leaves. Designating the

wheel as B and the pinion into

which the wheel gears as c,

Figure 1, the formula for the

problem reads as follows :

— = number of turns of the pinion. Figure 1

A PRACTICAL COURSE IN HOROLOGY

Substituting — by their numerical values we have :
c

— = 6 turns of the pinion to one of the wheel.
12

Calculating the number of turns of a complete train.
Most watch trains contain five wheels, as shown in Figure 2.
These include the main-spring barrel and the escape wheel,
and all five mobiles are usually named as follows :

B = barrel or first wheel
C = center or second wheel
T = third wheel
F = fourth wheel
E = escape wheel

The pinions are as follows :

c = center or second pinion
t = third pinion
f = fourth pinion
e = escape pinion

Figure 2. Main train.

WHEEL WORK 17

It will be noted that the second pinion is in reality the
first pinion, but for convenience it is given the same name
as the wheel of which it is a part and referred to as the
center or second pinion.

As an example of a modern train we shall select the num-
ber of teeth of wheels and leaves of pinions as shown below.
(Remember, the barrel, center, third, fourth and escape
wheels are indicated by the capital letters B, C, T, F and E,
and the center, third, fourth and escape pinions by the small
letters c, t,f and e.)

B 72

C _80_
T 75 _ ^

1 lo"^

F 80

Multiplying the numbers 6, 8, 7j^, and 10 together we
get 3,600, which represents the number of turns of the escape
wheel to one of the barrel.

Dividing 3,600 by 6 (turns of center wheel to one of the
barrel) we get 600. This number (600) represents the turns
made by the escape wheel in one hour, since the center wheel
carries the minute hand and, of course, makes one turn in an
hour. Henceforth all calculations will be based on one turn
of the center wheel, and on this basis we submit the follow-
ing formula, in which, as above, the capital letters C, T, F
indicate the wheels and the small letters t, f, e indicate the
pinions.

18 A PRACTICAL COURSE IN HOROLOGY

CTF 80 X 75 X 80

= = 600 turns of the escape wheel.

tfe 10 X 10 X 8
The fourth wheel in all watches designed to register
seconds is so constructed that a second hand is fitted to the
extremity of a long pivot extending through a hole in the
dial. It follows, therefore, that according to the accepted
plan for the division of time, the fourth wheel must make
60 turns to one of the center wheel. Therefore :

CT 80 X 75

= = 60 turns of the fourth wheel.

tf 10 X 10

However, this is not necessary in watches where there is
no second hand, and in many wrist watches the fourth wheel
makes more than 60 turns to one of the center wheel, as we
shall soon see.

Calculating the number of beats. The escape wheel in
most watches contains 15 teeth and delivers twice as many
impulses to the balance, since each tooth delivers two im-
pulses, first to the receiving pallet and later to the discharging
pallet. Letting E indicate the escape wheel and e the escape
pinion, the formula reads :

CTF2E

= number of beats per hour.

tfe ^

Substituting the numerical values we have :

80X75X80X2X15

= 18,000 beats per hour.

10 X 10 X 8

Fast and slow trains. Not all watches make 18,000 beats

per hour. Some of the older watches make 16,200 and 14,400

beats per hour, whereas some newer American wrist watches

make 19,800 and 21,600 beats per hour. There are other

trains of varying beats per hour, particularly among those

WHEEL WORK 19

of Swiss manufacture. Below are shown two examples of
slow trains :

80 X 75 X 72 X 2 X 15

== 16,200 beats per hour.

10X10X8

80 X 75 X 80 X 2 X 15

= 14,400 beats per hour.

10X10X10

As already stated, watches without second hands may have
any system of mobiles wherein the fourth wheel will not
register seconds. Below are listed three trains of this type,
all of which are fast trains used in small wrist watches.

54X 50X48X2 X 15

6X6X6
64X66X60X2X15

18,000 beats per hour.

= 19,800 beats per hour.

8X8X6

42 X 42 X 35 X 35 X 2 X 12

= 21,600 beats per hour.

7X7X7X7

The last train listed is unique in that the train has 6 wheels
with an escape wheel containing 12 teeth.

Comments on the fast trains. In the preceding para-
graphs we discussed the slow and fast trains that have been
in general use at various periods. We shall now consider
further the modern fast trains used to some extent in very
small wrist watches.

The watch train making 18,000 beats per hour has been
accepted as a sort of standard for many years. However,
very small wrist watches making beats of various numbers
per hour in excess of this figure are becoming more in evi-
dence and the increased use of such trains would lead one
to inquire into the reason for their popularity.

20 A PRACTICAL COURSE IN HOROLOGY

The reason lies in the fact that 18,000-beat trains, when
applied to very small wrist watches, have a tendency to set
on the locking. This is due to the fact that a light balance
and a weak balance spring have not the necessary force to
effect a satisfactory unlocking of the escapement. To over-
come this fault of setting on the locking, it is necessary to
reduce the drop lock to the very minimum. However, since
it is difficult to expect the same precision of adjustment in
small watches as is possible in larger watches, the fast trains
offer a solution. The fast trains make the unlocking easier
due to two factors: (1) the more rapid motion of the bal-
ance and (2) the increased strength of the balance spring.
In this way the new fast trains permit a greater tolerance
with regard to the extent of the lock, resulting in a reduced
tendency to set on the locking. Better timekeeping results.

There is another point in favor of the new fast trains.
Every horologist has observed how often the coils of the
balance spring get caught in the regulator pins or get badly
tangled due to a severe jolt. The stronger springs fitted to
fast train watches do much to eliminate the difficulty or, at
least, to lessen its frequent occurrence.

Calculating the number of teeth and leaves of missing

mobiles. We now wish to determine the number of teeth

of wheels and leaves of pinions that are missing. Let F

indicate the missing fourth wheel of the following train :

80 X 75 X F X 2 X 15

— —— ^ = 18,000

10X10X8

225F = 18,000

F = 80

According to the above solution the fourth wheel contains
80 teeth.

WHEEL WORK 21

In the next problem, the third pinion is mission. Letting t

indicate the mission pinion, the equation reads as follows :

80 X 75 X 80 X 2 X 15

— ——- — = 18,000

t X 10 X 8

180,000

= 18,000

180,000= 18,000 1

18,000 1 = 180,000

t == 10 leaves of the third pinion

Let us suppose that the complete fourth wheel and pinion
of a wrist watch are missing in which F is the wheel and /
is the pinion. The equation reads :

54 X 50 X F X 2 X 15

6xf X6

2,250 F

- = 18,000

= 18,000

f
F 18,000 8

f 2,250 1

The result shows that the fourth wheel should have 8
times as many teeth as the fourth pinion has leaves. Desig-
nating / by 6, 7, 8, or 10 leaves, we may obtain any of the
following solutions.

48 56 64 80

"6" '7 '¥ 'To

Any of the solutions given above may be used; however,
judging from the numbers of teeth and leaves of the train as
a whole, the most suitable numbers would be :

F 48

22 A PRACTICAL COURSE IN HOROLOGY

There are times when the ratios come out with a fractional

numerator. For example we wish to determine the number

of teeth and leaves of a missing third wheel and pinion in

which T is the wheel and t is the pinion.

54 X T X 48 X 2 X 15

-— = 18,000

tX6x6

2,160T

= 18,000

T 18,000 syi

T^ 2,160 "T"

The only numbers that may be tried without producing a

fractional number of teeth are :

75 50

— and —

9 6

In this example the most suitable numbers would be :
T 50

T~6~

There are other times when we arrive at the answer im-
mediately, as in the case of determining the number of the
teeth of the escape wheel, E, and the leaves of the escape
pinion, e.

90 X 80 X 80 X 2E

12 X 10 X e

9,600E

= 18,000

E 18,000 15

= 18.000

e 9,600 8

The answer shows that the escape wheel contains 15 teeth
and the escape pinion 8 leaves.

WHEEL WORK 23

Calculating a new train. We now wish to determine the

number of teeth for the wheels and number of leaves for the

pinions of a new train. Let us suppose the watch is to be a

small baguette making 21,600 beats per hour. Using the same

letters as before to indicate the wheels and pinions the

formula reads as follows :

CTF2E

= 21,600

tfe

We may decide in advance the number of leaves for the
pinions and the number of teeth for the escape wheel. For
the pinions t, f, and e we shall use 6 leaves each. The escape
wheel will have 1 5 teeth. The equation now reads :

CTF2X 15

21,600

6X6X6

21,000 X 6 X 6 X 6

CTF =

2X15

CTF= 155,520

The combined product of CTF is 155,520. In order to
obtain the quantities desired, it is necessary to resolve this
number into its prime factors and to form these factors into
three groups which will represent the numbers for the teeth
of the wheels C, T, and F.

2)155520 2) 9720 3) 405

2) 777Q() 2) 4860

2) 38880 2) 2430

2) 19440 3) 1215

5
Factoring, we find that CTF = 155,520 = 2^ X 3^ X 5.

135

24 A PRACTICAL COURSE IN HOROLOGY

These factors may be formed into groups of various com-
binations, but the most suitable arrangement for a watch
train would be as follows :

C = 5X3X22 = 60

T = 33 X 2 = 54

F = 3 X 2^ = 48

Thus we show the complete train.
60 X 54 X 48 X 2 X 15

6X6X6

= 21,600 beats per hour.

The Barrel and Its Mainspring

The barrel and its mainspring are important factors in the
design of a watch. The ratio between the barrel and center
pinion must show a definite relationship to the length and
strength of the mainspring and must be determined with
mathematical exactness if the watch is to perform satis-
factorily and run a required number of hours.

Calculating the number of hours a given watch will run.
In order to determine the number of hours a watch will run
we must first find the number of turns of the center wheel
to one of the barrel. Letting B indicate the barrel and c the
center pinion, the formula reads :

— = number of turns of center wheel to one of the barrel.
c

Using a numerical example, we have :

84 ^

— = 7 turns of center wheel.
12

Knowing that the center wheel makes one turn in an hour,
it follows that the barrel makes one turn in 7 hours. Next

WHEEL WORK 25

we must determine the number of turns necessary to com-
pletely wind the spring. A trial has shown that it takes 5^
turns to wind the spring. The number of hours the watch
will run is found by multiplying 7 by 5^, thus :
7X5^== 3Sj4 hours the watch will run.

The mainspring should run the watch not less than 32
hours; 36 to 40 hours is better; in fact, some of the finest
watches will run 45 hours and more.

Calculating the correct thickness of the mainspring.
We experience no difficulty in fitting mainsprings to standard
makes of watches, for all we have to do is to select the spring
as catalogued and graded by the particular manufacturer.
There are times, however, when an old watch or one of
unfamiliar make needs a new spring and, if we suspect that
the old spring is not the correct one, we need to apply a bit
of mathematical calculation to determine the correct thick-
ness of the spring. In such cases the following method is
suggested :

1. Divide the teeth of the barrel by the leaves of the center
pinion in order to determine the number of hours taken for
one turn of the barrel.

2. Determine the number of turns necessary to wind the
spring by dividing 36 (hours of running) by the number of
hours consumed in one turn of the barrel.

3. Measure the inside diameter of the barrel and divide
by 12.5.

4. Divide the above quotient by the number of turns neces-
sary to wind the spring. The result is the thickness of the
spring.

Suppose for example that the barrel, B, has 78 teeth and

26 A PRACTICAL COURSE IN HOROLOGY

the center pinion, c, has 12 leaves. The inside diameter of
the barrel is 12 millimeters.

1. — = — = 6.5 X 1 =6.5 hours,
c 12

2. = 5.5 turns to wind sprmg.

6.5

3. = .96

12.5

•96 ...

4. = .17 mm., thickness of the spnng.

5.5

It should be understood that the answers are only ap-
proximate. In some cases, as in a fine, 21- or 23- jewel watch,
a weaker spring may be needed, whereas a 7- jewel watch may
require a stronger spring.

Calculating the correct length of the mainspring. The
proper length of the spring need not be calculated in so many
inches. Instead, we may state that the spring should occupy
one half of the area between the inside wall of the barrel
and the periphery of the arbor. If the spring does this, the
length is correct.

Observe that we say area and not space. The term space
would lead one to infer that radial measurements are in-
tended, which would be incorrect, inasmuch as a spring
wound up would occupy more radial space than that of a
spring run down. Thus, the statement often found in older
books on horology to the effect that we allow one third of
the space for the arbor, one third for the space, and one third
for the spring is slightly in error and indicative of a spring
a few coils too long.

The correct rule should read :

WHEEL WORK

One third of the space is occupied by the barrel arbor and
one half of the remaining area is covered by the spring.

This is clearly shown in Figure 3. It will be observed
that the radial distance occupied by the spring is less than

Figure 3. Barrel showing correct space for mainspring.

that given to space. However the area of both spring and
space are equal and it remains the same under all conditions,
whether the spring is wound completely, partially let down,
or completely let down.

Now wind the spring in the barrel and if the spring oc-
cupies more than one half of the area, break off the outer
end and rewind in the barrel. When the correct area is
determined the hook is affixed. A spring of the proper length
contains from eleven to thirteen coils. More than the re-

28 A PRACTICAL COURSE IN HOROLOGY

quired number of coils only tends to increase friction and
shorten the number of hours of running of the watch.

The Dial Train

The cannon pinion, minute wheel, minute pinion, and hour
wheel make up the dial train. Referring to Figure 4, the

Figure 4. Dial train.

dial train is arranged as follows. The cannon pinion, c, gears
into the minute wheel, M. The minute pinion, m, to which
the minute wheel is attached, gears into the hour wheel, H.
This latter wheel fits freely over the cannon pinion. The
minute hand, of course, makes 12 turns to one of the hour
hand. The formula for the dial train, therefore, reads as
follows :
HM

= 12

Let us suppose we wish to determine the number of teeth
of wheels and the number of leaves of pinions of a new dial
train. We shall decide on 12 leaves for the cannon pinion, c,

WHEEL WORK 29

and 10 leaves for the minute pinion, m. Letting H indicate

the hour wheel and M the minute wheel, the equation reads :

HM HM

= 12

cm 12 X 10

HM=12X 12X 10; HM = 1,440

Factoring in the manner as heretofore explained, we find
that:

HM = 25 X 32 X 5
Combining these factors to form two groups representing
the wheels H and M we have :

H = 23 X 5 = 40; M = 22 X 32 = 36

Thus the complete train reads as follows :

HM 40 X 36

= — -- — =12

cm 12 X 10

Various types of dial trains. Dial trains may be grouped
into three types: (1) the regular, (2) the reverse, and
(3) the irregular. In the regular the ratio of cannon pinion
to minute wheel is 3 to 1 and the ratio of the minute pinion
to the hour wheel is 4 to 1.

M_36_ H_l^_4

"c"~T2~ ' m"~10"""

In the reverse the ratio of the cannon pinion to minute
wheel is 4 to 1 and the ratio of the minute pinion to hour
wheel is 3 to 1.

M_32_ H_?l_^

T~'%~ ' m~"8"~

The irregular does not belong to either of the above types,
as shown by the following example :

HM 35X48 48X26 32X45

cm 14 X 10 13 X 8 12 X 10

30 A PRACTICAL COURSE IN HOROLOGY

Thus the dial train may be formed by working out various
combinations, the only requirement being that the combined
ratios equal 12. The most used, however, is the 3 to 1 —
4 to 1 type referred to as the regular.

Calculating the teeth and leaves of missing mobiles.

Let us suppose that the cannon pinion is missing from the
following train in which c represents the cannon pinion :
54X32

CX12
i^ = 12

= 12

144 = 12c
12c = 144
c = 12 leaves of the cannon pinion.

We now wish to find the number of teeth of a missing
hour wheel in which H represents the wheel.
H X 25 _ ^^
10 X 10

« = 12
4

H = 48 teeth of the hour wheel.

Let us suppose that a complete minute wheel and pinion
are missing in which M is the wheel and m is the pinion.
48 XM

14 X m
24 M

= 12

7m
M 12X7_ 7 _3^

m"~~24 2 ~~ 1

WHEEL WORK 31

The result shows that the minute wheel must have 3^^
times as many teeth as the minute pinion has leaves. Thus
M_28_35_42_49
m~" 8"~10~12~~l4

Any of the above solutions may be used as the following
equations will show.

48X28 48X35 48x42 48x49

14 X8 14 X 10 14 X 12 14 X 14

= 12

Problems

1. What is meant by train ?

2. How many trains has the ordinary watch ? Name them.

3. Name the wheels of the average main train.

4. How many turns does the escape wheel make to one of the
center wheel in an 18,000-beats-per-hour train? How many turns
in a 19,800-beats-per-hour-train?

5. What are the arguments in favor of the fast trains for very-
small wrist watches ?

6. What portion of the space between the barrel and the arbor
should the mainspring occupy ?

7. What is the correct thickness of the mainspring if the barrel
has 80 teeth with an inside diameter of 13 millimeters and the center
pinion has 10 leaves?

S. Name the parts of the dial train.

9. How many turns does the hour wheel make in 12 hours ?
10. Name the various types of dial trains. Explain their difference.

Part I

CHAPTER TWO

Gearing

Terminology

Gearing constitutes a system of wheels and pinions
whose circumferences are covered with teeth so that the
teeth of the wheel act upon the leaves of a pinion. The
function is in reality as a system of levers in which a longer
lever presses on a short one until one lever ceases to press and
another lever comes into action.

Gearing is a rather technical subject and it Is well first to
study Figure 5. The terms given in the illustration are
defined as follows:

Pitch circle: a circle concentric with the circumference of
a toothed wheel and cutting its teeth at such a distance from
their points as to touch the corresponding circle of the pinion
working with it, and having with that circle a common
velocity, as in a rolling contact.

Pitch diameter : the diameter of the pitch circle.

Full diameter: the diameter from point to point of the
teeth.

Distance of centers: the distance measured on a straight
line from center to center between the wheel and pinion.

GEARING 33

Line of centers: a line drawn from center to center of
any wheel and pinion.

Circular pitch: the pitch circle divided into as many
spaces as there are teeth on the wheel or pinion.

Diametrical pitch: the diameter of the pitch circle
divided into as many spaces as there are teeth on wheel or
pinicn.

ADDENDUM
DEDENDUM

Figure 5.

Addendum: the portion of the tooth, either of wheel or
pinion, outside of the pitch circle.

Dedendum: the portion of the tooth of either wheel or
pinion inside of the pitch circle.

Driver : the mobile that forces the other along.
Driven: the mobile that is being forced along by the
driver.

Principles of Gearing
The addenda. In Figure 6 is shown a portion of a gircle,
A, representing the pitch circle of a wheel. Rolling on this
portion of a circle is another circle, B, the diameter of which

A PRACTICAL COURSE IN HOROLOGY

Figure 6. Formation of the epicycloidal curve.

equals half the pitch diameter of the pinion. If a pencil
point were fixed at the lowest point of the circumference of
the smaller circle (the generating circle) and then rolled on
the larger circle without slipping, a curve would be traced
along the path of the line C in the direction of the arrow.
The curve thus formed is called the epicycloid and determines
the shape of the addenda of the wheel teeth.

The dedenda. The dedenda of the pinion leaves is formed
by the same generating circle but not in the same manner.
The smallef circle is rolled inside and along the pitch circle
of the pinion. However, instead
of a curved line a radial line is
formed as shown by the line A
in Figure 7. A circle rolling
within a circle is called a hypo-
cycloid and determines the shape
of the dedenda of the pinion
leaves. Thus when a wheel and
pinion are made in conformance
with the above principles of de-

Figure 7.

GEARING 35

sign, a smooth and constant force is delivered to the pinion.

Width of tooth. It has been observed that the generating
circle forms one side of the tooth. The question now arises
as to how to determine the width of the tooth. This is
determined by dividing 360 (degrees in any circle) by the
number of teeth in the wheel. This gives us in degrees the
width of one tooth and one space, generally referred to as
the circular pitch.

360
Thus = 4.5 degrees circular pitch.

80 teeth

The width of the tooth is equal to one half of the circular

pitch ; the other half is, of course, equal to the space.

4.5 degrees

Therefore = 2.25

of which 2.25 degrees is the width of the tooth and 2.25

degrees the width of the space.

Now, by placing the generating circle, E, Figure 6, with
the pencil point directly below and 2.25 degrees to the right
of curve C, it will be in position to trace out the other side
of the tooth by simply rolling the circle to the left as shown
by the dotted line D. The intersection of the two curves,
C and D, forms the point of the tooth.

The pitch diameter. We now wish to determine the pitch
diameter of a wheel and pinion, the center distance being
known.* The first procedure is to determine the diametrical

* The center distance is determined by means of a depthing tool,
an instrument with adjustable male centers that can be adjusted to
the holes in the plate and convenienth' measured with a micrometer
or Boley gauge.

36 A PRACTICAL COURSE IN HOROLOGY

pitch, the formula for which reads as follows :

center distance X 2 . , . ,

: — r-. — = diametrical pitch.

teeth of wheel -j- leaves of pinion

The diametrical pitch is now multiplied by the number of
teeth in the wheel in order to determine the pitch diameter
of the wheel, and in like manner the diametrical pitch is
multiplied by the number of leaves in the pinion to determine
the pitch diameter of the pinion.

For example, the center distance is 8.5 millimeters ; the
wheel has 80 teeth; the pinion has 10 leaves. Substituting
the numerical values for the above formula, we have :

= .1888 diametrical pitch.

80+10 ^

Continuing the problem we find that :

.1888 X 80 = 15.104 mm. pitch diameter of the wheel.

.1888 X 10 = 1.888 mm. pitch diameter of the pinion.

„ ^ 15.104+1.888 ^ ,^, , ,.

Proof: = 8.496 mm., the distance between

centers.

The full diameter. The height of the addenda is a varying
quantity depending on the ratio of the wheel to the pinion,
but the production of theoretically correct gears or even
knowing when they exist is not possible with the equipment
available to the practising horologist. The usual practice is
to add 2.5 diametrical pitches to the pitch diameter of the
wheel and 1.25 to the pitch diameter of the pinion.* Experi-

* There is one exception to the above statement For the dial
train where the pinions drive the wheels and the wheels drive the
pinions, as in the case of setting the watch to time, the addenda is
figured as 2 for both wheels and pinions.

GEARING 37

ence has shown that the above figures are best for all practical

purposes.

We found that the wheel has for its pitch diameter 15.104
millimeters and the pinion 1.888 millimeters. The diametrical
pitch multiplied by 2.5 gives us the height of the addenda
for the wheel :

.1888 X 2.5 == .47
Adding this to the pitch diameter of the wheel, we have :

15.104 + .47 = 15.57 mm. full diameter of the wheel.

Now, figuring the pinion we have :
.1888 X 1.25 = .236
1.888 + .236 = 2.12 mm. full diameter of the pinion.

We may, however, figure the full diameters with a lot less
work by adding 2.5 or 1.25 (addenda) to the number of the
teeth or leaves. For example :

(80 4- 2.5) .1888 == 15.57 mm. full diameter of the wheel.

(10+1.25) .1888 = 2.12 mm. full diameter of the pinion.

Circular pitch. It will be noted that the definition for
circular pitch reads somewhat like the definition for diamet-
rical pitch. The difference is : circular pitch is the division
of the circumference of a circle (the pitch circle), whereas
the diametrical pitch is the division of the diameter of a circle
(the pitch diameter). In both cases the number of teeth or
leaves is the divisor.

We must know the actual width of tooth and space in order
to select a cutter to make a wheel. Herein lies the importance
of calculating the circular pitch. To attain this we make
use of the following formula:

pitch diameter X 3.1416

teeth or leaves

= one circular pitch

38 A PRACTICAL COURSE IN HOROLOGY

Substituting the numerical values :
15.1 X 3.1416

= .592 mm. circular pitch

The proportion of tooth or leaf to space is usually :

for the wheels : one half of the circular pitch.

for the pinions, 10 leaves or less : one third of the circular

pitch,
for the pinions, 12 leaves or more : two fifths of the

circular pitch.

Now, continuing with the above example to determine the
width of the tooth of the wheel and the leaf of the pinion,
we find that
.592

.592

= .296 mm., the width of the tooth, and
= .197 mm., the width of the leaf.

Problems

1, What is the epicycloid? the hypocycloid?

2. Define diametrical pitch.

2. Calculate the diametrical pitch for the following:
center distance — 7.5 millimeters,
wheel — 75 teeth,
pinion — 10 leaves.

4. Calculate the pitch diameter for the above wheel and
pinion. Calculate the full diameters.

5. Define circular pitch.

6. Calculate the circular pitch, using the specifications given
in Problem J.

Part I

CHAPTER THREE

The Lever Escapement

Time and experience have demonstrated the superiority
of the lever escapement over all other types for portable
timepieces. In fact, the several other types have now
become obsolete. Since its introduction by Thomas Mudge
in 1750 the lever escapement has been the object of much
experimentation and study. It finally was developed into the
state of perfection we see it today only after a number of
unique and fantastic variations were tried and discarded.

Terminology
The several parts of the escapement are defined as follows :

ESCAPE WHEEL
The escape wheel is that part of an escapement that delivers
impulse to the balance through the medium of a pallet fork.

Ratchet-tooth wheel: the name given to the English
type escape wheel, which has pointed teeth.

Club-tooth wheel: that type of escape wheel which has
a lifting face at the end of the teeth. Impulse face: the lifting
plane of a club-tooth wheel. Locking face: the slanting face
of the teeth on which the pallets lock. Toe: the intersection
of the locking face and the impulse face of a club tooth.

40 A PRACTICAL COURSE IN HOROLOGY

Heel: the intersection of the impulse face and the letting-off
corner of a club tooth.

PALLET FORK

The pallet fork is that part of an escapement that, by
means of pallet jewels, receives impulse from the escape
wheel and delivers impulse to the balance.

Pallets: the name given to the metal body to which the
lever is attached. The term includes the pallet jewels.

Lever: a metal piece attached to the pallets that carries
impulse to the balance. The fork occupies the extreme end
of the lever.

Fork slot; a notch cut into the fork for the reception of
the roller jewel.

Horns: the circular sides of the fork that lead to the
fork slot.

• Receiving pallet : that pallet stone over which a tooth of
the escape wheel slides in order to enter between the pallet
stones.

Discharging pallet : that pallet stone over which a tooth
of the escape wheel slides in order to leave from between
the pallets. Impulse face: the lifting plane of the pallet stone.
Letting-off corner: the extreme end of the impulse face of a
pallet stone where the tooth of the escape wheel lets ofiF.
Locking face: the face of a pallet stone on which a tooth locks.

ROLLER TABLE

The roller table is the circular disk that carries the roller
jewel.

THE LEVER ESCAPEMENT

Crescent: a circular notch in the edge of the roller table
for the reception of the guard pin or fingeV.

Single roller: a roller action comprising a single metal
disk.

Double roller : a roller action comprising two metal disks,
the larger disk carrying the roller jewel and a smaller disk
in which a crescent is cut.

Roller jewel or jewel pin: a flattened jewel that is
inserted in the roller table.

BANKING PINS
Banking pins are pins that arrest or limit the angular

motion of the lever.

Equidistant, Circular and S emit an gent al Pallets

There are three types of pallet arrangements used in the
lever escapement: the equidistant, the circular and the
semitangental.

Equidistant pallets. In the equidistant the locking faces
of the pallets are an equal distance from the pallet center,

Figure 8. Equidistant pallets.

A PRACTICAL COURSE IN HOROLOGY

as shown in Figure 8. As a result the lifting action on the
discharging pallet takes place too far from the point of
tangency, necessitating the need for a greater lifting angle
on that stone. Although the unlocking is performed under
favorable conditions, the lifting action is not, being unequal
in its distribution and unequal also as to the pressure of the
tooth on the pallets. This escapement calls for exceptional
accuracy in its construction.

c f

Figure 9. Circular pallets.

Circular pallets. Circular pallets have the central portion
of the pallets' lifting faces an equal distance from the pallet
center as shown by the lines CA and DA in Figure 9. One
half of the width of the pallets is placed on each side of these
lines, which requires that the locking faces stand at an un-
equal distance from the pallet center, causing an unequal and
increased unlocking resistance. However, the action of lift
is more favorable.

Semitangental pallets. Setting on the locking is a com-
mon fault in small wrist watches; hence a light lock is

THE LEVER ESCAPEMENT

Figure 10. Semitangental pallets.

essential to good performance. With regard to this factor
the semitangental escapement, a development of recent years,
has found favor with many manufacturers because it lessens
the resistance to unlocking, a natural fault with the circular,
and at the same time minimizes the lifting error of the
equidistant. Figure 10 shows the semitangental escapement.
It will be observed that the unlocking on the receiving pallet
takes place on the tangent EB at the intersection of CA.
A slightly increased center distance results, since the line
CA is placed 31 degrees to the left of the center line BA.
The action of the discharging pallet is similar to that of an
escapement with circular pallets. The designers claim that
the unlocking and lifting actions are more nearly equally
divided than any escapement that has yet been conceived.

Number of Teeth in Escape Wheel

Although nearly all watches have an escape wheel of 15
teeth we may make them with 12, 14, 16, or practically any

M A PRACTICAL COURSE IN HOROLOGY

number in this range. The first lever escapement as made
by Thomas Mudge had an escape wheel of 20 teeth. The
larger numbers result in a rather clumsy action and are
therefore not satisfactory. Escapements using a wheel of
12 or 14 teeth are used today in some wrist watches and have
certain advantages. For example, there is more clearance
between the pallet arm and the escape wheel. The smaller
number of teeth permit the use of wider pallet jewels and
a wider lifting plane for the teeth. Also the actual measure-
ment of 1^2 degrees of locking is greater because of the
increased distance between the locking comers of the pallets
and the pallet center.

Wheel and Pallet Action

The lever escapement has two distinct and separate func-
tions: (1) the action of the wheel and pallets and (2) that
of the fork and roller. These functions we shall now con-
sider. The first, which has to do with the wheel and pallets,
is divided into three actions. They are :

1 the locking

2 the draw

3 impulse or lift

The locking. The locking Is the overlapping of a tooth
on a pallet when the lever rests against the bank. This locking
is necessary for the reason that if one tooth lets of? a pallet
and another tooth falls directly on the impulse face of the
opposite pallet, there is a recoil of the lever toward the
balance, causing a frictional contact between the guard pin
and roller table. Contact in this manner would immediately
stop the watch. To avoid this a certain amount of lock Is

THE LEVER ESCAPEMENT 45

necessary. However, it should be as little as possible con-
sistent with the proper safety in action.

The draw. In portable timepieces the lock in itself is not
sufficient to insure safety in action. It is necessary to create
an action whereby the pallets are drawn into the wheel. This
is effected by inclining the locking faces of the pallets to the
right of the lines CA and DA as shown in Figure 10. This
slanting should be as little as possible, yet enough to overcome
the friction of the tooth on the surface of the pallets, for
it can readily be seen that the combined action of lock and
draw makes a resistance to the motion of the balance and
has an important relationship to the position and isochronal
rating of a watch.

It is generally conceded that 12 degrees draw on each
pallet is satisfactory. However, because of the circular
motion of the pallets the draw is changing continually, being
strongest on the receiving pallet at the point of unlocking,
whereas on the discharging pallet it is weakest at the point
of unlocking. The nature of this action leads one to assume
that a greater angle for draw should be given to the receiving
pallet ; in fact, we find some authorities designing the escape-
ment with 13 to 15 degrees draw on the receiving pallet.
This would make the draw more nearly equal at the point
of banking where it is most needed.

The lift. In well-designed escapements of the club-tooth
variety the actual lift is 8^ degrees, being divided between
tooth and pallets in varying proportions. Adding 1^^ degrees
for the lock, the total angular motion of the lever becomes
10 degrees.

Width of pallet and tooth. It will be observed that a

46 A PRACTICAL COURSE IN HOROLOGY

wide pallet requires a narrow tooth ; likewise a narrow pallet
should be associated with a wide tooth if the drop* is to be
kept to the minimum. In this connection several pallet and
tooth combinations are listed below :

CIRCULAR PALLETS

PALLET TOOTH

Lift Width Lift Width

1 6 degrees 7 degrees 2J^ degrees 3^ degrees

2 5J4 •' 654 " 354 " 4>^ "
5 5" 6 " 3^ " Ay2 "
^ 4 " 5 " 4H " 5^ "

EQUIDISTANT PALLETS

Lift Width Lift Width

5 6 degrees 7 degrees 2y2 degrees 3 5^ degrees

6 5H " 6 " 3 " 454 "

7 5^ " 5M " 3 " 4^ "

Attention should be given to this fact : in the equidistant
pallets the lift on the tooth should be less than the Hft on the
pallets. Circular and equidistant pallets are not always
interchangeable. However, there are exceptions, as in the
case of Number 1 and Number 5, which are alike and have
been interchanged by Grossman. Number 4 is suitable only
for circular pallets, especially so since the narrow pallets
perform the act of locking nearer to the lines CA and DA,
as shown in Figure 9. Numbers 5 and 6 would be suitable
also for the semitangental escapement.

* Drop is a term used to indicate the free motion of the escape
wheel after one tooth lets off a pallet and another tooth locks on
the opposite pallet. This factor is of more concern in practical
benchwork and is treated fully in Part II, Chapter Six.

THE LEVER ESCAPEMENT 47

The Fork and Roller Action

Unlocking and impulse actions. The relation between
the fork and roller jewel as a mechanical action may be
divided into two distinct and separate functions: (1) the
unlocking of the pallets and (2) the impulse to the balance.
One action is the reverse of the other. The unlocking action
takes place as a result of power derived from the balance
and spring, while the impulse to the balance receives its
energy from the force delivered to the escape wheel by means
of the main spring and the train. In the unlocking action,
a short lever (roller- jewel radius) acts on a longer one (the
lever fork). A short roller-jewel radius must be associated
with a long lever which involves a large angle of contact as
shown by the angle ABC in Figure 11. On the other hand,
a long roller- jewel radius must be associated with a short
lever which results in a small angle of contact as shown by
the angle ABC in Figure 12. In other words, the shorter
the roller- jewel radius the larger is the angle of contact, and
the smaller the angle of contact the longer is the roller- jewel
radius.

A 4-to-l roller action. In Figure 11 the relation of the
roller jewel to the fork slot is 4 to 1 ; that is, the lever moves
10 degrees and the roller jewel remains in contact with the
fork slot for a space of 40 degrees. The action of unlocking
begins quite some distance from the line of centers because
of the short roller-jewel radius. However, a short roller-
jewel radius results in a much safer action and an easier
unlocking of the pallets.

A 3-to-l roller action. In Figure 12 the relation of the
roller jewel to the fork slot is 3 to 1. Although the unlocking

A PRACTICAL COURSE IN HOROLOGY

action is more difficult, the impulse to the balance is more
energetic when it does occur. The 30-degree contact of the
roller jewel with the fork slot is in accord with the theory
that the lever should be as highly detached as possible for
finer results in timing.

Figure 11. Fork and roller ac-
tion— single roller.

Figure 12. Fork and roller ac-
tion— double roller.

The more delicate safety action of a long roller-jewel
radius is not a problem in high-grade watches, for in such
work we naturally expect a mechanical action that is more
exact in its function. It follows, therefore, that a long roller-
jewel radius which involves a small angle of contact is one

THE LEVER ESCAPEMENT 49

of the important factors to be considered in fine escapement
design. Of course, in the cheaper work and in small wrist
watches it is not always practical.

A wide roller jewel. The wide roller jewel results in a
particularly satisfactory action of unlocking by taking place
near the line of centers. However, the impulse to the balance
is not so favorable. A wide roller jewel is inseparable from
a long roller- jewel radius, for such a principle of design
results in a very delicate action. A wide roller jewel makes
the safety action much less delicate.

A narrow roller jewel. A wide roller jewel, if associated
with a short roller- jewel radius, causes an unfavorable im-
pulse to the balance too far from the line of centers, and the
"uphill" circular path of the roller jewel along the side of
the fork slot during the impulse to the balance results in
considerable friction. Also, on account of the greater angle
at which the roller jewel stands to the slot when the impulse
takes place, the drop of the fork against the jewel will amount
to more than its shake in the slot, if measured when standing
on the line of centers. Thus a narrow roller jewel is better
adapted for a short roller- jewel radius, and should always
be associated with a single-roller escapement.

Single-roller escapement. In the single-roller escapement,
Figure 11, it is required that the roller table be as small as
possible to preserve the safety action. Furthermore, friction
between the guard pin and the circumference of the roller
table would be considerably increased on a table roller that
was larger than necessary. It is difficult to reduce the angular
motion of the lever to less than 10 degrees, and any relation
of fork to roller jewel less than a 3^ to 1 is not practical.

so A PRACTICAL COURSE IN HOROLOGY

Double-roller escapement. ' We have learned,' that ii^
order to favor the impulse to the balance we require a long
roller-jewel radius, and for the safety action a short radius.
This is the reason for the passing of the single-roller escape-
ment in favor of the double-roller type, Figure 12, for in the
latter type we have two rollers, one for each action.

The size of the safety roller is of no great importance.
For the sake of soundness in action its radius should not be
less than one half the radius of the roller jewel. The smaller
the safety roller the sooner will the crescent approach the
guard finger; and, likewise, the longer the roller-jewel
radius the later will the roller jewel enter the fork slot. It
follows, therefore, that the greater the difference between
the respective radii of the roller jewel and the safety roller
the longer must be the horns of the fork. The width of the
roller jewel also plays a part in the length of the horns, for
with any increase in the width of the jewel, the horns may
be made proportionately shorter.

The crescent. The circular notch in the roller is called
the crescent and should be wide and deep enough so that it
will be impossible for the guard finger to touch any part of
it. If made too wide, longer horns on the fork would be
required to preserve the safety action.

The width of the crescent in the double-roller escapement
is greater than in the single-roller type, for the reason that
the guard finger, due to its increased length, has a larger
space to cover for its safety action, or, stating the function
in another way : it could be said that the velocity of the guard
finger has increased, whereas the velocity of the safety roller
has decreased.

THE LEVER ESCAPEMENT 51

Problems

1. How do the equidistant pallets differ from the circular pallets?

2. Compare the above with the semi-tangental escapement.

3. What is meant by locking?

4. Define draw. How does draw differ on the receiving pallet
as compared with the draw on the discharging pallet ?

5. How many degrees are generally intended for the lifting
action ?

6. Explain the difference between a 4 to 1 and 3 to 1 roller action.

7. Which of the above actions is preferred? Why?

8. Describe the single and double escapements. State the advan-
tages and disadvantages of each type.

9. What is the name of the small roller that makes up the double-
roller escapement ?

10. What is the purpose of the crescent?

Part I
CHAPTER FOUR

The Controlling Mechanism

The balance and balance spring are the most vital parts
of a watch and may be properly called the controlling mechan-
ism. Years ago the principal difficulty in maintaining accu-
rate time was the temperature error, since the compensating
balance was unknown prior to 1769. Often the error was as
much as four or five minutes in twenty-four hours. The
variation is due to three conditions : ( 1 ) the expansion and
contraction of the metal in the balance, (2) changes in the
length of the balance spring, and (3) variation of the elastic
force of the spring. The variation of the elastic force of
the spring is the most important factor; in fact, Ferdinand
Berthoud has estimated that 82 per cent of the error is due
to the variation of the elastic force.

Experimental demonstration. A simple experiment to
prove the correctness of the above statement may be tried

Figure 13.
52

THE CONTROLLING MECHANISM S3

if desired. The materials needed are a sheet of brass about
4 or 5 inches long and 2 inches wide, a piece of brass wire,
and an old mainspring. Drill two holes, A and B, Figure 13,
about yi inch apart and insert two brass pins and rivet
securely. Straighten the outer portion of the mainspring
and place the end between the pins, the extreme end being
placed under pin A and over pin B, while the coiled portion
passes beyond the index C. The coil beyond the index will
serve as a weight. Now, with the aid of an alcohol lamp and
blowpipe heat the brass plate. It will be observed that
the spring, after becoming heated slightly, has deflected
noticeably, as may be seen by noting the position of the
spring on the index. Upon cooling it will return to its orig-
inal position.

Temperature Error of the Balance and Balance

Spring

The compensating balance. The compensating balance,
or bimetallic balance, as it is sometimes called. Figure 14,
was designed to overcome the errors resulting from the use
of the solid balance. This balance is constructed by brazing
together brass and steel for the rim of the balance. The
brass is on the outside and constitutes about three fifths of
the total thickness.

Heat causes the metals in the balance assembly to expand ;
the arms become longer and, as the brass expands more than
the steel, the loose ends of the rim curve inward toward the
center. Cold causes the loose ends to move outward away
from the center, while the arms become shorter. This is
clearly shown in Figure 15. It will be further observed that
the loose ends remain reasonably circular during temperature

A PRACTICAL COURSE IN HOROLOGY

Figure 14. Compensating balance,

changes, but the radii of the curves change, their centers
being at the balance center O for the normal temperature and
shifting along the arms AA for the high and low tempera-
tures. However, the points BB, about 60 degrees from the
arms, remain at a fixed distance from the balance center, and
it is at these points that alterations for the purpose of timing
should be made.

Middle-temperature error. It is evident from the above
analysis that we can adjust the balance screws in such a
manner as to compensate for the expansion and contraction
of the balance alone and maintain a constant mean diameter.
This, however, would not take care of the lengthening and
shortening of the balance spring nor for the changes in the

THE CONTROLLING MECHANISM

Figure 15.

elastic force. To compensate for the effects of temperature
on the spring, it is necessary to add extra weight to the loose
ends of the rim. This results in a temperature error, between
the extremes of heat and cold, known as the middle-tempera-
ture error due to the fact that the balance does not com-
pensate equally for changes in the elastic force and for
changes in the length of the spring. This is shown by in-
sufficient compensation (weights not moving in near enough
toward the center of the balance) in the higher temperatures ;
and too great a compensation (weights moving too far away
from the center) in the lower temperatures. The result is
a higher rate in the normal temperature, usually from two
to six seconds in twenty-four hours, depending on the grade
of the watch.

A PRACTICAL COURSE IN HOROLOGY

Figure 16.

This factor is clearly shown in Figure 16. The line AB
indicates the uniform loss in the rate due to a rising tempera-
ture on the balance spring only. To exactly offset this rate
by some means of compensating the error, we must produce
the opposite effect indicated by the line CD. The theoretical
results would be a constant mean rate along the line EF.

2ec

onJs

4
3
2
1
0

'0° S0° 60° 70° 80° &i

Figure 17. Middle-temperature error.

THE CONTROLLING MECHANISM

However, since the actual compensating effect of the balance
alone lies along the dotted line GH, the actual rate is similar
to that shown in Figure 17.

There is no way of rectifying this condition except to use
a nickle-steel alloy called Invar, the perfection of which has
been realized only in recent years. This remarkable metal
expands and contracts only a very little for changes in
temperature, and in using it instead of the ordinary steel
the middle-temperature error is considerably reduced.

Still more recently another type of alloy made of iron,
nickle-chromium, and tungsten and known as Elinvar has

Figure 18. Solid balance.

58 A PRACTICAL COURSE IN HOROLOGY

been developed for the use of balance springs. Elinvar main-
tains a constant elasticity and is, therefore, used in connec-
tion with a solid, single-metal balance as shown in Figure 18.
The metal has the further advantages of being nonrusting
and only slightly subject to magnetism. Also there can be
no middle-temperature error.

The Balance Spring

There are two forms of balance springs in general use.
These are the flat spring, Figure 19, and the Breguet, Figure
20. The former has the stud fixed to the same plane as the
body of the spring, with the result that the vibrations take
place in an eccentric manner. The latter, referred to as
Breguet, has a portion of the outer coil raised above and
over the body of the spring. The original Breguet spring
was created by Abraham Louis Breguet (1747-1823), famous
French horologist, but the spring by him bore no resemblance

Figure 19. Flat balance spring.

THE CONTROLLING MECHANISM

-YA

Figure 20. Overcoil balance spring.

to the theoretical terminals as applied to modern watches.
The terminals as we now find them are based on the founda-
tion laid down by M. Phillips, French mining engineer, and
have placed the art of adjusting on a scientific basis.

Theoretical terminals. The value of the theoretically
corrected terminals lies in their capacity to correct position
and isochronal errors by eliminating the eccentric wanderings
of the center of gravity that are everpresent in the flat spring.
The outer and inner terminals designed by L. Lossier perform
this function very satisfactorily. These are shown in Figures
20 and 21.

In the outer terminal, Figure 20, note that the overcoil is
composed of portions of two circles. The outer coil tends

A PRACTICAL COURSE IN HOROLOGY

inward at A and forms an arc
of 180 degrees to B, which is
planted at a point equal to .67
of the radius R. The overcoil
continues for another 83 de-
grees, the radius of which is
CO. To this must be added
the amount necessary to reach
through the stud.

Figure 21 shows the Lossier
inner terminal. The elements
of the curve are the same as
Figure 21. Theoretical inner terminal, the outer terminal and are
clearly shown in the illustration. Although the overcoil is
applied to practically all watches today, the theoretically
corrected inner terminal is not, its use being confined only
to the very finest watches. The reason no doubt lies in the
fact that the Lossier inner terminal is difficult to make and
harder to adjust and, unless the curve is 100 per cent perfect,
it is no better than a true terminal of the ordinary type.
Careful poising of the collet is also necessary.

Problems

1. Describe the compensating balance.

2. What is the middle-temperature error?

3. What type of balance springs are used with solid, single-metal
balance wheels ?

4. How does the Breguet spring differ from the flat spring?

5. What are the advantages of the theoretical terminals?

6. Draw a sketch showing the proportions of the theoretical
curves, both outside and inside.

PART II
PRACTICAL REPAIRING

Part II

CHAPTER ONE

Train Problems

General observations in gearing. Attention should
always be paid to the action of the outgoing tooth, noting
that its point is not used and that the incoming tooth takes
up its load without drop. The incoming tooth should begin
pressing on the pinion leaf as near as possible to the line of
centers, allowing for the fact that this is a varying quantity
depending on the number of leaves in the pinion. The best
possible actions for pinions of ten, eight, and six leaves are
shown in Figures 1, 2, and 3. Note that in Figure 1 the action

begins on the line of centers, in
Figure 2 slightly before the line
of centers, and in Figure 3 quite
some distance from the line of
centers.

Gearing may be found defec-
tive with regard to several fac-
tors. Below are listed the most
common.

Pinion too large
Pinion too small
Depthing too deep
Figure 1. Ten-leaf pinion. Depthing too shallow

TRAIN PROBLEMS

Figure 2. Eight-leaf pinion. Figure 3. Six-leaf pinion.

Let us now examine the errors in the order listed above.

Pinion too large. Referring to Figure 4 it will be ob-
served that the incoming tooth butts into the end of the pinion

Figure 4. Pinion too large. Figure 5. Pinion too small.

A PRACTICAL COURSE IN HOROLOGY

leaf, thereby stopping the watch. If the error is slight the
action can be improved by enlarging the wheel, in which case
the outgoing tooth will propel its leaf farther, resulting in a
sufficient clearance between the incoming leaf and the in-
coming tooth.

Pinion too small. A pinion too small, Figure 5, results
in a waste of power and much noise, since the outgoing tooth
propels the leaf even to the point of slipping off before the
incoming tooth has started to press on a leaf. Hence, if much
too small, the incoming tooth will fall with a click on the
leaf. Wear is considerable and an error of this kind should
never be allowed to pass without correction. The fitting of a
larger pinion is the only satisfactory solution.

Figure 6. Depthing too deep. Figure 7. Depthing too shallow.

Depthing too deep. In a depthing too deep, Figure 6,
the outgoing tooth continues its action too long, most likely
with the point affected by excessive wear and a loss of power.

TRAIN PROBLEMS 65

Depthing too shallow. A shallow depth, Figure 7, is
always unsatisfactory. Note that the outgoing tooth cannot
propel its leaf far enough and the incoming tooth presses on
a leaf before the line of centers. In a very shallow depth a
butting action usually results. Enlarging the wheel some-
times puts the depth in a passable condition.

Worn pinions. If a watch stops in the train and starts
with the slightest movement it is well to look at the pinions
and see if they are worn. Modern wrist watches frequently
use a six-leaf escape pinion, which when well designed gives
good service, but has the disadvantage of about 18 degrees
engaging in friction. Wear is as a result quite pronounced
and in time will cause trouble. Often the fourth wheel can
be raised or lowered so that the wheel drives the pinion on
the previously unused portion.

Repairing a Train

Stretching a train wheel. The enlarging of train wheels
is a job that comes occasionally to the horologist, for it
happens that some watches are not right when they leave the
factory. It is, of course, better to fit a new wheel, but there
are times, particularly if the watch is not an expensive one,
that enlarging the old wheel is permissible. The fact remains
that if the wheel was imperfect to begin with and wx make it
serve its purpose and the watch runs better, we feel justified
in the act.

The staking tool may be used to stretch the wheel. It is
preferred that we work on the lower side of the wheel, so
that the markings made by the flat punch will not be visible
when the watch is assembled. We first select a hole in the

66 A PRACTICAL COURSE IN HOROLOGY

die to loosely fit either the staff of the pinion or the entire
pinion as the case may require. The die is so adjusted that
the rim of the wheel comes under the punch. It frequently
happens that the part of the rim we wish to work on comes
over, or partly over, another hole in the die. When this
happens we may select a larger hole and plug it with pith.
This will hold a staff in position while stretching the rim
over a smooth and solid part of the die.

To stretch a wheel we use a flat- face punch of such width
as to completely cover the rim including the teeth. Tap the
punch gently a succession of blows and at the same time turn
the wheel slowly. This stretches the rim satisfactorily and
with scarcely any visible punch marks, except where the arms
are crossed.

To keep the wheel as nearly round as possible, it is neces-
sary to stretch the arms, as we should not rely on the cutter
of the rounding-up tool to bring the wheel circular. Stretch-
ing of the arms is done separately — ^that is, after the enlarging
of the rim and teeth — and great care should be exercised
so as not to overdo the stretching.

The rounding-up tool. Having finished the stretching
we are ready to use the rounding-up tool. First, select one
of the brass beds the diameter of which is sufficient to sup-
port the wheel and at the same time be perfectly free of the
cutter. Next, select a cutter that exactly fits the space be-
tween the teeth. Occasionally, we find wheels the teeth of
which are too wide. In such cases a wider cutter is required
in order to reduce the teeth to a width that will satisfactorily
gear into the pinion. Having selected the cutter and placed
it in the tool, we center it in line with the guide, a knifelike

TRAIN PROBLEMS 67

piece provided for the purpose of making certain that the
teeth will be cut perfectly upright. Tha wheel to be rounded-
up is now placed between centers and so adjusted that the
wheel turns freely and the rim barely touches the brass bed.
Carefully advance the cutter so as to engage it with the
wheel. See also that the guide on the cutter is properly
centered within the space preceding the one that the cutter
occupies. Cut the wheel but proceed carefully. It is well to
try the wheel frequently in the watch so as to not overdo the
correction.

Problems

i. How many leaves are required of a pinion so that the action
begins on the line of centers ?

2. Name four defects that are sometimes found in gearing.

3. If a train is noisy what may be the reason?

4. Why is a shallow depth always unsatisfactory?

5. If a watch stops frequently in the train, what may be the
reason ?

Part II

CHAPTER TWO

Jeweling

The use of jewels as bearings for watches is, without
question, one of the most important achievements to the
attainment of precision timekeeping. Nicholas Facio, an
Italian residing in London, successfully applied jewels to
watches about the year 1704. The system used by Facio was
not the same as employed in making the jeweled bearing of
today. Instead of a hole piercing a jewel, a V-shaped depres-
sion was ground into the jewel. The pivot was pointed and
worked into the depressed jewel in much the same way as in
the present-day alarm clock. The Swiss were quick to
realize the advantages of jeweling and began experiments
which finally resulted in the making of jewels as we find
them today.

Bezel-Type Jeweling

Jeweling of the bezel type is a rather difficult task when
attempted by the usual hand methods. Yet the lathe attach-
ment intended for such work is practical only when a con-
siderable number of jewels are to be set. We shall, therefore,
describe only the hand method — a method which, after the
necessary experience, will satisfy the needs for all practical
purposes. Few tools are needed. These are a supply of drills,
the usual gravers, a jewel graver, Figure 8, and a burnisher,

Figure 9.

JEWELING

Side
Figure 8, Jewel graver.

Bottom

Figure 9. Jewel burnisher.

The procedure is as follows. Secure in the lathe a piece
of four-millimeter brass wire. Face off the end of the wire
and turn a small center. Select a drill slightly smaller than
the jewel to be used and bore a hole about five millimeters
deep. With the jewel graver enlarge the hole slightly to true
it up. Next turn a seat to fit the jewel. The depth should be
sufficient so that the jewel will lie slightly below the surface
of the wire. Next, cut a groove close to the opening for the
jewel with a long, pointed graver. The jewel is now to be
inserted but should first be moistened with a little oil to
keep it from falling out. Now rest the burnisher on the
T-rest; thrust the point of the burnisher in the groove,

A PRACTICAL COURSE IN HOROLOGY

Figure 10.

Figure 11. Bezel type jewels.

forcing the brass against the jewel, thereby holding the jewel
in place. A jewel thus set is perfectly secure and the brass
may be faced off level to the jewel if desired. Figure 10
clearly shows the work as described above. The dotted lines in
the figure show the manner in which the setting is turned to
fit the watch plate. End shake is tried before the wire is cut
off. This being satisfactory the wire is cut off, turned to
the proper thickness, and stripped out with a sapphire jewel
stripper or a highly polished graver. The face is polished
by sliding the setting on an agate polishing stone or burnish
file previously prepared with a Number- 1 buff.

Three styles of bezel-type jewels are used in watches and
are shown in Figure 11. The method for setting all styles
is the same.

Friction Jeweling
Friction jeweling of watches is a simple and quick method
of inserting a jewel in a plate, bridge, or setting by means of
friction. Swiss manufacturers started using this system in
1920, and since then more and more manufacturers, both
Swiss and American, have adopted this method.

JEWELING

In fitting a friction jewel, the first procedure is to deter-
mine the depth the jewel is to be set to give the proper end
shake. This is accomplished by using a machine especially
made for the purpose, of which there are many varieties on
the market. First, rest the pusher on the broken jewel and
adjust the metric screw near the top of the machine so that
the new jewel may be forced in to the same depth as the
broken one had been. Second, push out the broken jewel
and ream out the hole with the smallest reamer that will cut
away enough metal to give to the plate a clean, straight hole.
Third, select the proper jewel, the outside diameter of which
is 1/100 of a millimeter larger than the hole in the plate.
Remove the burr left by the reamer with the wheel counter-
sinks and push the jewel in place.

Replacing a friction jewel. To replace a jewel in a
watch that had a friction jewel in it before, it is necessary
only to measure the size of the hole in the plate. This may
be done by inserting the reamers or using a special gauge
that is available for the purpose. Having determined the
hole size, select the jewel required and push it in to the proper
depth.

Figure 12. Friction Jewels.

n A PRACTICAL COURSE IN HOROLOGY

Fitting jewel in removable setting. If we wish to fit a
jewel in a setting that may be removed from a plate, as in
the case of a balance or cap jewel, we need special tools to
hold the setting securely while reaming out the hole. There
are various types of tools on the market all of which are
used in connection with the friction- jeweling machine.

Several types of friction jewels are shown in Figure 12.

Problems

1. What are the tools needed in fitting bezel-type jewels?

2. Name the styles of bezel-type jewels used in watches.

3. How do friction-type jewels differ from bezel-type jewels?

4. How do you determine the depth a friction-type jewel is to be
set?

Part II

CHAPTER THREE

Making a Balance Staff

Many horologists consider the making of a balance staff
a difficult task. We find workmen doing almost anything to
a watch to avoid the necessity of making a staff. Balance
bridges are bent up or down. Unsightly graver marks are
found on plates and bridges. The balance arms are sometimes
bent out of line in an attempt to permit the balance to clear
the various parts, and pivots are often ground too short.
However, the making of a staff is not difficult if the repair-
man would go about learning the art the same as with any
other performance requiring skill. No one ever learned to
play a musical instrument in a few lessons or ever became
an expert engraver in a few months.

Preliminary Notes on Staff Making

Gravers for turning staffs. Three gravers of the styles
shown in Figure 13 are needed for staff work. A is for
general use, suitable also for square shoulders and the
cylindrical portion of cone pivots. B has a rounded point
for turning the cone portion of the cone pivot. C has the
point flattened and is used for turning the lower end of the
staff prior to cutting it off.

A PRACTICAL COURSE IN HOROLOGY

ABC

Figure 13. Gravers for turning balance staff.
Sharpening the gravers. Sharpening the gravers does not
seem to be given the attention by the average horologist that
it should. We have seen v^orkmen trying to cut a square
shoulder pivot with a graver having a point like a wire nail.
Naturally their work was unsuccessful, yet these workmen
did not reflect on the fact that possibly the graver was at
fault. The graver must have a sharp point. Even the best
gravers cannot retain a keen edge very long while cutting
tempered steel. Therefore keep a sharpening stone handy and
make frequent use of it. Some workmen use an emory or
carborundum wheel to grind gravers. This should never be
done, as the point of the graver is frequently softened and
this point is the most important part. Instead we use two
stones, a soft Arkansas stone and a hard Arkansas stone.
The soft stone is for rapid cutting and the hard stone is for
the final finishing. After grinding the face, sHde the sides on
the stone so as to produce a flat and smooth cutting surface.

Making pivots for practice. The beginner should prac-
tice making square shoulder and cone pivots before attempt-

MAKING A BALANCE STAFF

ing the making of a staff. It will be found that the larger
pieces of pivot wire are excellent for the purpose, as the
wire is hardened and tempered, ready for use. The length of
the cylindrical portion of a cone pivot is twice the diameter.
The length of a square shoulder pivot is three times the
diameter.

Measuring for the staff. Now, returning to our problem
of making a staff, the first act is to take the necessary meas-
urements, preferably from the watch, for the reason that the
old staff may not be correct. The well-known Boley gauge
serves the purpose very well, since it reads both ways,
between the calipers and from the end of the foot. See that
the balance bridge lies flat with the lower plate. Now remove
the cap jewels. Make certain that the hole jewels are
securely pushed in place. For the full length of the staff A,
Figure 14, measure from the side of the lower hole jewel to

Figure 14.

76 A PRACTICAL COURSE IN HOROLOGY

the side of the upper hole jewel. The height of the seat for
the roller table B is found by measuring the distance from th^
side of the lower hole jewel to the top of the lever, adding
enough for clearance and the thickness of the roller table.
In like manner the distance for the balance seat C is measured
from the side of the lower hole jewel to the top of the pallet
bridge, adding for the necessary clearance. The length of
the upper end of the staft D is found by subtracting the
length of the lower end to balance seat C from the full
length A.

Preparing the steel. Preparing the steel wire for the staff
is next in order. Select a piece of steel, the diameter of which
will be a little larger than the largest part of the staff when
finished. Heat over a gas flame to a cherry red and plunge
quickly into water. This should be done in a rather dark
place so as to see better the degree of heat, for if the light
is too bright, one is apt to overheat the steel and thereby
ruin it.

The wire is now too hard to turn and we must therefore
draw the temper. The wire must be made white in order to
blue it. This is done in the lathe by holding a piece of fine
emory paper against the wire.

The tempering is done by drawing the wire through the
flame of an alcohol lamp; or, better still, lay the wire on a
curved sheet of copper, keeping the wire rolling while being
held above the lamp. A full blue color is satisfactory for
staffs.

Turning the staff. Tighten the wire securely in the lathe,
having extended the wire from the chuck sufHciently to
include the full length of the staff and about two millimeters

MAKING A BALANCE STAFF

Figure 15. Pivot polisher.

additional. With the hand graver turn the upper end — first
turning the balance seat to length from the end and almost
to size, say .05 of a millimeter of the finished diameter. Next
turn the collet axis, followed by turning the remaining end
nearly to the size of the hole in the roller table.

Turning a conical pivot. Turn the cylindrical portion of
the pivot almost to size. Next, using the graver with the
point slightly rounded, turn the cone, bringing it down to
meet the cylindrical portion of the pivot. This is followed
by cutting the slope between the collet axis and the cone of
the pivot. The turning of the upper end of the staff is now
completed and we are ready for the pivot polisher and the
preparation of the grinding mediums.

The pivot polisher. The pivot polisher, Figure 15, is a
very neat little instrument for grinding and polishing. The
superiority of the instrument over any hand method is un-
questionable ; it does the work in a factory-like manner and
polishes the pivot most beautifully. It is to be regretted that
the pivot polisher is not more generally used.

A PRACTICAL COURSE IN HOROLOGY

However, the successful use of the pivot polisher depends
on the proper preparation of the laps and we shall digress
for a moment to consider the method by which they are
made — how to keep them in good condition, and their vari-
ous uses. We need laps of cast iron, bell metal, and box-
wood, and the material may be purchased from hardware
stores and material houses.

Cast-iron and bell-metal laps. Laps made of cast iron
and bell metal are used principally for grinding. The mate-
rial may be purchased in rods of various diameters. For most
needs of the horologists, rods of about ^ inch in diameter
will suffice. Having selected the materail, saw off a piece
about ^2 inch long and bore a hole in it of such size that a
reamer of the required taper may be used to enlarge the hole.
The hole is reamed out with a reamer of the same taper as
the taper chuck shown in Figure 16. The blank is placed in
the taper chuck for turning with the slide rest. Laps of
various shapes are needed. Those most used are shown in
Figure 17.

Figure 16. Taper chuck.

Figure 17. Laps for pivot polisher.

MAKING A BALANCE STAFF 79

Having turned the laps, the face and side must be filed to
enable the lap to hold the grinding or polishing medium.
Filing, it is admitted, has a tendency to destroy its truth, yet
this is necessary in order to do good work. Experience has
shown that it is possible to file a lap many times and still
run practically true. File the lap by laying it face down on a
Number 4 or 5 file. Holding the lap between the thumb and
fingers, slide the lap along the cutting teeth of the file about
an inch or so. The lap should then be turned partly around
and another stroke made. This crosses the lines, providing
a suitable surface for the embedding of the abrasive medium.
Next, prepare the side of the lap. This is done by drawing
the lap on the file in a direction parallel to the hole in the
lap. Continue in this manner until the entire circumference
is filed.

Boxwood laps. Boxwood laps are used for putting a high
polish on steel. We may use the slide rest in turning the
boxwood in the same manner as we did in preparing the
cast-iron and bell-metal laps. It is important that the grain
of the wood run parallel with the hole in the lap so that the
end of the grain touches the work. The lap is filed on the
face only with a Number 0 file.

Preparing the grinding material. Now that the laps are
ready, the preparation of the grinding medium is next in
order. We proceed as follows :

Apply a small quantity of oilstone powder to the first com-
partment of the three-compartment polishing block. Add a
little watch oil and mix with a small knife or spatula until a
thin paste is produced. Place the pivot polisher on the lathe
and adjust the lap spindle so that it stands at the same height
from the bed as the lathe spindle. Adjust the index at the

A PRACTICAL COURSE IN HOROLOGY

base of the polisher to }i degree, so as to give the staff a
slight taper toward the end when the grinding takes place.
With the belting so fitted that the grinding surface of the lap
and the surface of the staff rotate in opposite directions to
each other, feed the lap up to the work by means of the hard-
rubber knob at the rear. Apply thinly but evenly the oilstone
paste to the lap and grind the balance seat. Continue the
grinding until the balance just fits the seat. Be sure the
undercutting is deep enough so that the corner of the lap
does not touch the staff. If this is the case the balance can
be riveted true and flat when the staff is finished. Next,
grind the collet axis to size ; after which, grind that portion
of the staff between the cone of the pivot and the collet seat.

Grinding and polishing a cone pivot. Clean the work of
the grinding material with a piece of pith previously dipped
in benzole and hold it against the staff. Finish cleaning with
a dry piece of pith. For the cone pivot, the polisher is set
with the spindle at right angles to the pivot with the index
set at 0 degree. It is further adjusted so that the center of

the lap stands above the pivot
as shown in Figure 18. This
reduces the straight portion of
the pivot perfectly cylindrical
and forms the cone at the same
time. The shape of the cone
can be varied by raising or
lowering the spindle of the
polisher.

Instead of the cast-iron lap

Figure 18. Method of polishing
conical pivot.

we now use a bell-metal lap
and a grinding medium of cro-

MAKING A BALANCE STAFF 81

cus. Reduce the pivot, frequently trying the jewel until it
fits rather closely. Now remove the bell-metal lap and fit in
its place a boxwood lap and polish the pivot, using a paste of
diamantine and oil previously prepared on the top section of
the polishing block. Continue the polishing until the pivot
fits the jewel freely.

The slope between the collet axis and the cone of the pivot
may now be smoothed further by holding a jasper slip in
the hand. It is then polished with a boxwood slip and
diamantine.

Turning hub and roughing out lower end of staff. Now
that the upper end of the staff is finished we next turn with
the hand graver a long slope from the balance seat to the
lower end of the stafif. With the pivot polisher set at the
necessary angle, grind the slope for most of its length, using
oilstone powder and oil. Smooth further with jasper slip
and polish with boxwood lap, diamantine, and oil.

Next, cut out roughly the lower part with a graver slightly
flattened at the point. Leave the hub a little longer than
the finished staff is to be and the roller axis a little larger.

Cut off the staff and measure the overall length. The staff
is, of course, a little too long. Place the collet axis in a split
chuck and stone off the lower end until the staff is worked
down to the correct length. Remove the staff and remeasure
the overall length between the calipers of the Boley gauge
as many times as is required, taking only a little oflf the end
at a time so that the staff will not be made too short. This is
positively the most exact method of bringing a staff to the
correct length. Any other method, such as measuring back
from a predetermined amount of extra length and cutting
off after the work is cemented, is likely to result in errors.

A PRACTICAL COURSE IN HOROLOGY

Figure 19, Cement chuck.

The cement chuck. The staff having been brought to the
correct length we are now ready for the cement chuck. A
screw brass tightly screwed in a chuck must have a center
turned in it, deep enough to take the staff up to and in-
cluding the hub. Figure 19 shows the center with the staff
in place and properly secured in the cement.

Setting the staff in the cement chuck. Place a small
alcohol lamp under the cement brass. Heat sufficiently so
that the brass will melt the cement and fill the center, at the
same time running the lathe slowly. While the cement is
still soft, insert the staff with the thumb and first finger.
Again keeping the lathe in motion, reheat the brass until
the cement adheres to the staff. Holding the staff in place
with a hollowed-out piece of pegwood resting on the T rest,
continue running the lathe until the cement is slightly cooled.
Now, true up the staff while the cement is still soft by
resting the pointed end of the pegwood on the T rest and
bringing the pointed end against the roller axis. The T rest
must be placed at an angle so that we may first hold the
staff in place while allowing the cement to cool. The truing

MAKING A BALANCE STAFF 83

follows immediately by holding the pegwood in front of
and at a right angle to the staff.

Turning the lower end of the staff. Turn the lower end
of the staff, measuring from the end up to the roller seat, and
bring the hub to the correct length. Reduce the roller axis
to 0.1 of a millimeter of the correct diameter. Set the pivot
polisher in position with the index adjusted to ^ degree
taper and grind the roller axis until the roller table slides
on and wedges tight at a distance about two thirds the way
up to the hub. The pivot is next turned and polished as per
instructions already given.

The staff is now removed from the cement chuck. Fill the
boiling pan or test tube with alcohol and boil off cement.

Finishing the ends of the pivots. The ends of the pivots
have not yet been finished but it is well first to try the staff
in the watch and test for end shake, so we may know better
in what manner to proceed. Since the measurement for the
length was made without allowance for end shake, it usually
follows that a slight touching up of the ends of the pivots is
needed. To accomplish this, place the staff in a split chuck
and flatten each end slightly, using a hard Arkansas slip.
Polish further with a jasper slip and finally finish with a
hardened steel burnisher and round the corners slightly.

COMMON ERRORS IN STAFF MAKING

Beginners are very apt to overlook certain important
details in the turning of a staff. One of the most common
is the absence of a square shoulder for the roller seat, so
that the roller table will not lie flat to the full diameter
of the hub. Undercutting for this purpose is not objection-

84 A PRACTICAL COURSE IN HOROLOGY

able, for many watch factory staffs are made in this manner.
In like manner the balance wheel should fit the staff, that is,
flat to the full diameter of the hub.

The collet axis, the balance axis, and the roller axis should
show a definite taper so that the parts in question will fit
properly. This is easily attained with the pivot polisher, for
the instrument may be set to the desired taper. In using
the hand graver, the eye must be trained to recognize a
suitable taper.

The straight portion of the pivots should be cylindrical
and not tapering as we sometimes find them, and the cones
should be shaped alike. Again these conditions are made
possible with the pivot polisher and likewise more difficult
to attain without it. Satisfactory undercutting at the top of
the balance seat is also important.

Success will be realized if the beginner will pay strict
attention to detail, and it will be found that staff making is
not as difficult as some workmen would have you believe.

Problems

1. What gravers are needed in turning a staff?

2. How do you measure for a staff?

3. Explain in detail the hardening and tempering of the steel for
staff making,

4. Name the steps in turning the upper end of a staff in their
proper order.

5. How do you go about bringing the staff to the proper length ?

6. Must the staff run absolutely true in the cement chuck?

7. Name the points of particular importance in staff making.

Part II

CHAPTER FOUR

Pivoting

Success in pivoting depends largely on the quality of the
drills and in keeping the drills sharp. It is better to drill the
pinions without tempering, yet there are times when temper-
ing must be resorted to, and in such cases a small copper
wire shaped as shown in Figure 20 may be used. The leaves
of the pinion are held in a pin vise or parallel pliers to prevent
the heat from spreading to that part while the copper is
heated over an alcohol lamp. The end of the pinion is
tempered to a blue color.

The pinion may be centered in the split chuck if it runs
true; otherwise, the cement chuck must be used. Face off
the pivot to the square shoulder and turn a small center.
Place the pivot drill in a pin vise and drill a hole, which
should be a little deeper than the length of the average pivot.
Should the drill cease to cut, sharpen it immediately. A dull
drill is apt to burnish the bottom of the hole, presenting a
serious problem. Should this happen, flatten the end of the
drill, which usually results in the drill cutting again. The
hole having been drilled to a satisfactory depth, secure a
piece of pivot wire in a pin vise, and file sufficiently to just
start and hang in the hole in the pinion. Caution should be
exercised in filing the wire to show as little taper as possible.

A PRACTICAL COURSE IN HOROLOGY

Roll the wire as the filing takes place so as to leave the wire
round. Finish with the Arkansas slip. Next, force the wire
in the hole, cut off the wire with the cutting pliers, and tap
the plug with a small hammer, thereby forcing the plug
securely in the hole. The pivot is turned nearly to size with
the hand graver and finished to the proper size as already

III I I I I I I-T1

I I I I II umi

Figure 20. Copper wire in position for tempering pinion.

PIVOTING

explained. It is advisable to undercut the square shoulder
slightly before polishing in order to keep the corner sharp as
shown in Figure 21.

Figure 21.

Problems

1. What are the important points to remember in drilling a pinion
for repi voting ?

2. If a drill ceases to cut what may be done so that the drill will
cut again ?

3. What would be the most Hkely reason for a pivot working
loose when turning with the graver ?

Part II

CHAPTER FIVE

Fitting Balance Springs

In fitting balance springs the first procedure is to deter-
mine the number of vibrations of the balance per minute.
This may be found in any watch fitted with a second hand
by first dividing the number of teeth in the fourth wheel by
the number of leaves in the escape pinion ; then, by multiply-
ing the quotient by twice the number of teeth in the escape
wheel, we have ascertained the number of vibrations of the
balance per minute.

For example, the fourth wheel has 80 teeth; the escape
pinion has 8 leaves ; the escape wheel has 1 5 teeth.

80X30

= 300 vibrations of the balance per minute

Other examples are as follows :
72X30

64X30

= 270 vibrations per minute

= 240 vibrations per minute

In watches where the fourth wheel does not register
seconds we have to go back to the center wheel, as shown in
the following examples :

54 X 50 X 48 X 30

6X6X6

= 18,000 vibrations per hour

FITTING BALANCE SPRINGS

64 X 66 X 60 X 30
8X8X6

= 19,800 vibrations per hour

Fitting a Flat Spring

In fitting a flat spring to a watch, it is necessary to know
the proper size. This is determined by centering the spring
to the balance jewel as shown in Figure 22. The correct
radius is located on the first coil that stands inside the
inner regulator pin. Next, we desire to determine the ap-
proximate strength, which may be found by suspending the
balance and spring an inch or two above the bench. If the
distance between the point where the spring is held by the
tweezers and the rim of the balance is J^ inch, the spring
is approximately of the strength
desired. It does not matter if the
balance is large or small ; the
distance remains practically the
same. Of course, this does not
necessarily decide that a particu-
lar spring is to be used, but it does
eliminate all unnecessary attempts
at vibrating a spring that is posi-
tively unsuited for a balance.

After trying several springs,
finally selecting one that conforms
to the above test, the next pro-
cedure is to vibrate the spring.
This can be done by counting the
vibrations or by using an instru-
ment commonly known as a vi-
brator. Since nearly all watches Figure 22. Method for deter-

r . J '.1 mining correct size of flat

of recent years are made with a ^^^^^^ ^^^-^^^

90 A PRACTICAL COURSE IN HOROLOGY

300-beats-per-minute train, it is convenient to use the vibrator,
about which more will be written in the next paragraph. For
the other trains it will be necessary to count the vibrations,
using a watch of known accuracy. The counting is done with
every vibration that takes place in a clockwise direction ; that
is, the return vibration is not counted. The usual practice is
to suspend the balance and spring by means of tweezers sup-
ported in the lathe bed. The lower pivot of the balance staff
rests on the watch crystal. Thus, counting for one minute
there would be 150 counts for a 300-beat train, 135 counts
for a 270-beat train, etc.

The vibrator. Vibrators may be purchased from material
houses or may be made from an old balance and spring
properly timed. Material catalogs may be consulted if the
horologist desires to make his own instrument.

The method of using the vibrator is that of comparison —
that is, watching the balance spring to be vibrated and the
vibrator balance and noting whether or not both vibrate in
unison. We may slide the tweezers as much as a half a coil
in toward the center of the spring or out a quarter of a coil
toward the outside without altering the radius too much for
a satisfactory fitting of the spring.

Having found a spring that vibrates properly, break it off
one quarter of a coil beyond the vibrating point. Next the
spring is pinned in at the stud and the balance bridge with
balance and spring attached is placed in the watch for the
final timing.

Fitting the Breguet Spring
The fitting of the Breguet spring to a watch involves all
of the work of fitting a flat spring, plus the forming of the

FITTING BALANCE SPRINGS

Figure 23.

overcoil. The diameter need not be so exacting; however,
it should not be larger than one half of the diameter of the
balance, including half of the screws.

We shall assume that the watch in need of a new spring is
ideally suited to the Lossier terminal shown in Figure 23.
The procedure is as follows: Vibrate the spring in the flat
and break off the spring about two millimeters beyond the
vibrating point. The length of the active portion of the
overcoil must be determined and for this we must resort
to a bit of calculation. We desire to find the length of the
portion BC and AB (Figure 23). The following formula
gives us the length of BC.

Number 1 :

2 CO X 3.14 X S3

BC =

360

92 A PRACTICAL COURSE IN HOROLOGY

The radius CO multiplied by 2 gives us the diameter of
that part of the overcoil concentric to O. Hence the diameter
2 CO multiplied by 3.14 gives us the circumference, which in
turn is divided by 360 to determine the length of I degree.
Multiplying the length of 1 degree by 83 degrees gives us
the length of 5C.

We now desire to know the length of AB, which, when
added to BC, gives us the total length of the curve.

Number 2:

(AO + OB) 3.14

AB =

Since AO + OB equals the diameter and since one half of
the circumference is required, it is only necessary to multiply
AO + OB by 3.14 to give the circumference and divide by 2.

Now, taking a practical example, let us suppose that the
diameter of the spring is 8 millimeters. The radius would be
4 millimeters. According to the elements of the curve, CO
is .67 of the radius. Thus :

CO = 4 X .67 = 2.68 mm.
Substituting this value for CO in equation Number 1 we
may determine the angular distance for BC.

5.36 X 3.14 X 83

BC = = 3.88 mm.

360

Substituting again for equation Number 2 we learn the

length of AB.

^ (2.68 + 4)3.14 ^^^^

AB = -^ -^— ^ = 10.48 mm.

For the full length we add 3.88 and 10.48, giving us 14.36
millimeters.

FITTING BALANCE SPRINGS 93

We now grasp the spring at the point intended for the
regulator pins and slide the spring along a scale in order
to determine the actual length of the curve, allowing for
sufficient extra length for the spring to reach the stud. It
is permissible to make a slight mark at this point to indicate
the starting point of the curve.

Raising the overcoil. We are now ready to raise the
overcoil. This is accomplished with rather heavy tweezers.
Grasp the spring with the tweezers about 25 degrees from
that point where the inward bend starts. Hold the spring
tightly and press into a piece of softwood until the outer coil
takes on the shape shown in Figure 24. Turn the spring
over and, measuring 22 degrees from the first bend, bend
again until the outer coil lies level wnth the body of the spring,
Figure 25. Bring in the raised portion so that the overcoil
takes the form shown in Figure 23. To form the overcoil,
tweezers of many shapes are desirable. These are shown in
Figure 26. Alterations should be made gradually, being
careful not to bend the spring too much, for in so doing, the

Figure 24. First bend in forming overcoil.

Figure 25. First and second bends in forming overcoil.

A PRACTICAL COURSE IN HOROLOGY

Figure 26. Balance spring tweezers.

spring is liable to be considerably weakened and possibly
broken.

The Lossier curve is not adaptable to most small watches
for the reason that the curve is brought in nearer to the
central portion of the total mass than these watches will
permit. However, the above analysis may be used as a basis
for calculating other forms, since the length of the curve
does not vary very much. The forms more adaptable to
present-day small watches are shown in Figures 27 and 28.
In Figure 27 note that the radius of the overcoil along the
path of the regulator pins is three fourths of the radius of
the spring. The length of the curve is 180 degrees plus 20
degrees. Again in Figure 28 the portion of the overcoil at
the regulator pins is placed in a position nearly in line with
the full radius of the spring. In this type, the curve usually
takes the form shown in the illustration and the bend for
the overcoil starts at a point opposite the regulator pins.

Converting a Flat Spring into a Breguet
In order to obtain a closer position and isochronal rate in
a watch originally fitted w4th a flat spring, it is sometimes
desirable to make the spring over into a Breguet. In many

FITTING BALANCE SPRINGS

Figure 27.

Figure 28.

96 A PRACTICAL COURSE IN HOROLOGY

watches where there is room enough under the balance
bridge this can be done without much difficulty. The curve
suitable for a case of this kind is composed of quadrants
of two circles connected by a straight line as shown in
Figure 29. The radii of the circles A and B are equal to
about one half of the radius of the spring. Owing to the
spiral nature of the spring the radius of the circle 5 is a
little longer than that of circle A. The dotted line CC shows
the outer coil of the flat spring and the solid line DD shows
the same coil after it was formed into a Breguet. Figure 30
shows the same spring with the circles and dotted lines
removed. A clearer picture of its relative form is thus
realized. The only change to be made to the watch is to
shorten the regulator pins and raise the stud. The curve is

c \

Figure 29.

FITTING BALANCE SPRINGS 97

Figure 30.

theoretically correct and if properly executed it will be found
to be equal to any type of overcoil.

Eccentric Motion of the Spring
After a spring has been fitted to a watch an eccentric
vibration may be observed due to a faulty form of the over-
coil. The rules for correcting the eccentric motion are as
follows :

1. If the eccentric motion takes place opposite the regula-
tor pins (as in a flat spring), bring in part of the overcoil
toward the center of the spring.

2. If the eccentric motion takes place on the same side as
the regulator pins, move part of the overcoil hack into the
main body of the spring.

Truing Balance Springs

The attainment of successful balance-spring truing comes
only with continued practice and patience. It is one of those

A PRACTICAL COURSE IN HOROLOGY

accomplishments that are difficult to teach through the
printed page. For this reason we shall make only a few
general statements as to the manipulative operations.

In truing the round it is never necessary to manipulate the
spring beyond the first quarter of the inner coil, assuming
that the spring is true except for that portion which is likely
to be out as a result of pinning in at the collet. Figure 31
shows a spring divided into quarters. The sections are
referred to as first quarter, half, third quarter, and pinning
point. The spring may be wide at or near any of these points
and the procedure in truing consists of pushing or twisting
the spring in the desired directon. The dotted lines in Figure
32 show the manipulating required to bring the spring true
in the round.

Figure 31.

FITTING BALANCE SPRINGS

Figure 32. Dotted lines indicate manner in which the balance spring is

twisted to true in the round.

In truing the flat the spring is pushed down or raised up
at those points that are high or low.

Problems

1. Give the formula for determining the number of beats per
minute of a watch.

2. How do you determine the correct size of a flat spring for a
given watch ?

3. How do you determine the correct size of a Breguet spring?

4. What is the correct length of a Lossier outer curve if the
diameter of the balance spring is 6.5 millimeters ?

5. How do you go about forming the overcoil ?

6. How do you correct an eccentric motion of the spring that
takes place opposite the regulator pins ?

7. How do you correct an eccentric motion that takes place on
the same side as the pins ?

Part II

CHAPTER SIX

Escapement Adjusting

In this study of the lever escapement we are particularly
concerned with practical benchwork ; that is, the placing in
good order the escapement of a given watch. It is important,
however, that we have some understanding of the theoretical
principles involved and w^e shall indulge, therefore, in as
much theory and escapement design as is necessary to aid
in the execution of practical repair problems.

The importance of understanding the escapement cannot
be overestimated. Large pocket watches often function
quite satisfactorily with faulty escapements but with small
wrist watches it is very different. The escapement in small
ladies' watches must be practically perfect. Since the larger
per cent of the watches that are brought in for repair today
are wrist watches there is need for greater skill in escapement
work. Inadequate knowledge results only in an endless
amount of trouble with watches that stop persistently, though
perhaps only occasionally. Erratic rates, too, can be traced
to defective escapements,

Wheel and Pallet Action
The best way to obtain a practical understanding of the
escapement is to proceed step by step, studying the separate

100

ESCAPEMENT ADJUSTING 101

functions, after which the escapement action as a whole will
be analyzed. Our attention will first be directed to the prob-
lem of banking the escapement to the drop.

Banking to the drop. The term "banked to the drop"
means that the banking pins are turned in such a position
that a tooth of the escape wheel will slide past the letting-off
corner of a pallet, thereby permitting the lever to reach the
opposite banking pin.

In order to effect a banked-to-the-drop condition, it is
first necessary to turn in both banking pins. The impulse
face of one pallet will now show contact with the impulse
face of a tooth of the escape wheel, but owing to the fact
that the banking pins have been turned in, the tooth is unable
to pass the letting-off corner of the pallet. Let us assume that
the impulse faces of the receiving pallet and a tooth show
contact. Turn the banking pin, against which the lever now
rests, away from the line of centers slowly until a tooth
passes the letting-off corner of the receiving pallet. At this
instant another tooth will lock on the locking face of the dis-
charging pallet. Next, move the lever to the opposite banking
pin, resulting in a contact being shown between the impulse
faces of the discharging pallet and a tooth. Turn the banking
pin, against which the lever rests, away from the line of
centers until the tooth drops as already explained, and the
job of banking to the drop is completed.

DROP LOCK
The extent of the lock on the pallets after an escapement
has been banked to the drop is called drop lock. This lock
takes place the instant a tooth drops on the locking face
of a pallet.

102 A PRACTICAL COURSE IN HOROLOGY

In Swiss watches and some American wrist watches it is
not practical to bank the escapement to the drop because of
the fact that the banking pins are not supplied with eccentric
screws. In this case the usual practice is to slowly move the
lever until the escape tooth drops and at the same instant
cease moving the lever and take note of the extent of the
lock on the pallets. A slight additional motion of the lever
should be required before the lever will reach its bank, which
is, of course, beyond that of drop lock. The additional motion
is called slide and will be considered further in the later
portion of this chapter.

Correct drop lock. Drop lock is a varying quantity, de-
pending on the position of the pallet stones in the pallet
arm, but it should be as light as possible consistent with
proper safety in action. A drop lock of 1^^ or 2 degrees is
the amount usually adopted for pocket watches, whereas 2 or
3 degrees is allowed for wrist watches.

Altering the drop locks. If the drop locks are too light
or unsafe, a deeper lock can be had by moving out one or
both pallets. Likewise if the drop locks are too deep a lighter
lock can be had by moving in one or both pallets. It will
be observed that when one pallet is moved the lock is changed
on both pallets and any alteration of the pallets should always
be followed by rebanking to the drop.

Out of angle. The lever should move an equal distance
on either side of the line of centers. If the lever does not
move an equal distance the lever is said to be "out of angle."
If the drop locks are deep, out of angle can be corrected by
moving in the pallet on the side where the lever's angular
motion is shorter from the line of centers. If the drop locks

ESCAPEMENT ADJUSTING 103

are light, out of angle can be corrected by moving out the
pallet on the side where the lever's angular motion is longer
from the line of centers. If the drop locks are satisfactory,
out of angle (if slight) can be corrected by carefully bending
the lever as close as possible to the pallet staff. This can be
done by holding the lever with a small pair of pliers and
bending it with the thumb and first finger. And remember —
rebank the escapement to the drop after each alteration.

THE DRAW
The force that keeps the lever against its bank is called
"draw." It exists because of the inclination or slant of the
pallet's locking face and the shape of the tooth. It will be
observed in Figure 36 that the receiving pallet inclines in the
direction toward the pallet center and the discharging pallet
inclines in the same direction but away from the pallet
center, thus forming the angle for draw. A draw of 12
degrees is considered sufficient by most horologists.

Examining the draw. Take a watch oiler or similar small
tool and lift the lever away from its bank, but not far enough
to cause the escapement to unlock. Now release the lever,
and in so doing the lever will, if the escapement is correct,
return immediately to its bank. Try this again with the lever
against the opposite banking. The lever should at once return
to its bank.

The purpose of draw is to avoid unnecessary friction be-
tween the guard pin and the roller table. When the watch
receives a jolt the lever is thrown away from its bank and
the guard pin comes in contact with the roller table, but the
action of draw causes the lever to return at once to its bank.
Draw should be sufficient to effect the return of the lever
to its bank for all ordinary conditions. If the draw is exces-

104 A PRACTICAL COURSE IN HOROLOGY

sive, there will be an unnecessary recoil of the escape wheel,
causing a condition whereby too much of the force of the
balance and spring is used in unlocking the escapement. The
result is a waste of power and a shorter arc of motion of
the balance.

Altering the draw. As a rule want of draw is due to in-
sufficient angle of a pallet stone. In most cases the jewel may
be tilted in the pallet arm. If the jewel fits tightly, a thinner
jewel may be substituted or the slot may be widened to give
the jewel more angle.

THE DROP

Drop is defined as the free motion of the escape wheel at
the time when one tooth passes the letting-off corner of
a pallet and another tooth comes in contact with the locking
face of the opposite pallet. Drop may be also defined as the
distance a tooth of the escape wheel travels without doing
any work.

Examining the drop. With a tooth locked on the re-
ceiving pallet observe the space that separates the letting-off
corner of the discharging pallet from the heel of the tooth.
Now move the lever to the opposite banking pin, thereby
causing a tooth to lock on the discharging pallet. Next ob-
serve the space that separates the letting-off corner of the
receiving pallet from the heel of the tooth.

Of course, the drop should be equal, but we do not always
find it so. A small drop on the receiving pallet is called
"close outside," and a small drop on the discharging pallet
is called "close inside." These errors call for correction. If
close inside, tip one or both pallets away from the pallet
stafif. If close outside, tip one or both pallets toward the
pallet staff.

ESCAPEMENT ADJUSTING 105

Usually, moving one pallet is all that is necessary. The
question may arise as to which pallet to move. This would
depend largely on the condition of the draw, the drop locks,
and the angular motion of the lever from the line of centers,
for we shall soon learn that whenever a pallet is shifted for
any reason all of the above conditions are altered.

SHAKE
Shake is defined as that space separating the letting-ofif
corner of the pallet from the heel of a tooth when the
opposite pallet is locked at the lowest locking corner. Shake
is similar to drop except that the examination for shake
occurs at the time when the tooth is locked at the lowest
locking corner of the pallet. When moving the lever away
from its bank but not enough to unlock the escapement, it
will be observed that a slight recoil of the escape wheel has
taken place. This action lessens the space between the letting-
off corner of the pallet and the heel of the tooth, showing
that shake is always less than drop. If an escapement has no
shake the watch will stop.

THE LIFT
Modern escapements have a total lift of 83^ degrees. The
amount of lift separately on tooth and pallet is designed in
varying proportions in different makes of watches. The lift
in a club-tooth escapement is a very complicated action and
requires considerable study to understand it. It will be
noticed that the lift does not function exactly the same on
both pallets. On the receiving pallet the wheel moves up as
also does the pallet and the pallet's locking corner moves with
greater velocity than the letting-off comer. On the discharg-
ing pallet the condition is reversed and the wheel moves

106

A PRACTICAL COURSE IN HOROLOGY

Figure 33. Lift on receiving pallet

Figure 34. Lift on discharging pallet.

down while the pallet moves up. Also, the letting-off comer
of the discharging pallet moves with greater velocity than
the locking corner.

A good action between the wheel and pallets is shown in
Figures 33 and 34. Note that as the tooth leaves the locking
faces of the pallets, the toe of the tooth shows contact only
with the pallet's impulse faces. Contact in this manner con-
tinues completely across the pallets until finally the two
impulse faces meet nearly parallel, and after that the heel

ESCAPEMENT ADJUSTING 107

Figure 35. Curved pallet stones by A. Lange & Son.

of the tooth passes the letting-off corner of the pallets. The
heel of the tooth comes into action quicker on the discharging
pallet ; also there is a tendency for the tooth to move faster
along that portion of the lifting plane near the letting-off
corner. The opposite takes place on the receiving pallet;
that is, the action is faster at the start of the lift.

To obviate this difficulty, A. Lange and Sons some years
ago made watches with escapements so designed that the
receiving pallet had a convex lifting face, the discharging
pallet a concave lifting face, and the lifting faces on the teeth
were also curved (Figure 35). This system, it will be ob-
served, would cause the lifting action of the tooth to func-
tion at a more nearly constant velocity.

Loss in the lift. There is a definite amount of loss of
lift on the discharging pallet of the club-tooth escapement.
This is shown in Figure 36. 5C is a straight line but the
escape wheel describes the circle DD ; hence these lines must
deviate from each other. In order that 5^^ degrees of lifting
take place, a lifting angle of 6% degrees is required of the
discharging pallet. However, this loss of lift is a problem to
be reckoned with principally in the equidistant pallets. In
the circular pallets the loss of lift is very little for the reason

108

A PRACTICAL COURSE IN HOROLOGY

c
a

•3

ESCAPEMENT ADJUSTING

109

no A PRACTICAL COURSE IN HOROLOGY

that the tangents are planted mid-way between the locking
and letting-oif corners of the pallets (Figure 37). The loss
of lift in the semitangental escapement is also small, usually
amounting to about one-half degree (Figure 40).

The Fork and Roller Action
In this study of the lever escapement we have up till now
concerned ourselves with such factors as banking to the drop ;
drop lock, draw, drop, shake, and the lifting action. These,
it is observed, constitute the escape wheel and pallet action.
We are now ready to investigate the fork and roller action
which is quite a study in itself. There is, however, a definite
relationship between the two actions and the last mentioned
cannot be successfully studied without taking particular
account of the former. Hence the wheel and pallet action
in its entirety was outlined first and it is assumed that our
escapement is in satisfactory adjustment up to this point.

The lever's angular motion. We have stated that the
combined lift on the tooth and pallets is S}^ degrees. Adding
this to a drop lock of 1^ degrees, the total angular motion
of the lever becomes 10 degrees. The lever's angular motion
of 10 degrees should be all that is necessary for the roller
jewel to pass in and out of the fork satisfactorily without
catching. Now, placing the balance in the watch, we are
ready to try the tests for the safety locks. These are the
guard safety test and the corner safety test. Using a strong
eyeglass in making the tests, we proceed as follows.

SAFETY LOCK TESTS
Guard safety test. Rotate the balance so that the roller
jewel stands outside of the fork and with the first finger

ESCAPEMENT ADJUSTING

111

Figure 3S. Guard safety test

hold the balance in this position. Now, with a watch oiler or
similar small tool, lift the lever away from its bank, thereby
causing the guard finger to come in contact with the edge
of the safety roller as shown in Figure 38. With the lever
held in this position examine the remaining lock on the pallet.
This remaining lock is called a safety lock and it should

112 A PRACTICAL COURSE IN HOROLOGY

represent one half of drop lock or }i
degree of lock. The test should next
be tried on the opposite pallet and a
similar lock should be found.

Corner safety test. Starting with
the roller jewel in the fork slot, rotate
the balance slowly until such time that
one tooth passes the letting-off corner
of a pallet and another tooth comes in
contact with the locking face of the
opposite pallet. A slight additional
motion applied to the balance will bring
the roller jewel in a position opposite
to the slot corner. With the balance
held in this position, lift the lever away
Figure 39. from its bank, thereby causing the slot

Corner safety test. corner to come in contact with the
roller jewel as shown in Figure 39. With the lever held thus,
examine the remaining or safety lock. Try this test on the
opposite pallet and if the safety lock is the same on both
pallets the lever's angular motion from the line of centers is
practically equal.

The safety lock shown by the comer test should be the
same as the safety lock shown by the guard test ; that is, the
safety locks from both sources should show }i degree of
lock. Note — Although this is correct in theory it does not
always work out in practice. If the safety lock shown by the
corner test is less than the safety lock shown by the guard
test — that is, if the corner freedom is greater than the guard
freedom — no harm results, provided that the locking of both
pallets is safe on all of the teeth of the escape wheel. How-

ESCAPEMENT ADJUSTING 113

ever, if the corner freedom is less than the guard freedom,
the roller jewel is apt to catch on the tips of the horns of the
fork, causing the watch to stop.

The curve test. To test an escapement for the error stated
above, we use what is called the curve test. To apply this
test it is necessary to rotate the balance so that the roller
jewel stands completely past the horns of the fork. Next, lift
the lever away from its bank, thereby causing the guard
finger to come in contact with the safety roller and, while
the lever is held thus, turn the balance so that the roller jewel
will move toward the fork slot. If the roller jewel passes the
horns of the fork and enters the slot, the escapement is satis-
factory as far as this test is concerned. If the roller jewel
catches on the tips of the horns, a number of faulty condi-
tions could be present. The most common are: guard finger
too short, roller jewel advanced too far, or lever too long.

The drop locks being correct, it is now apparent that the
guard, corner, and curve tests aid in determining the correct
length of the lever. It is common practice in escapement
work to lengthen the lever by stretching it or to shorten it
by grinding the horns as the case may require. So bear in
mind that the condition of the drop locks is first taken ac-
count of and the fork and roller action afterward.

Slide
Up to this point in our discussion, the escapement has
been banked to the drop. The subject of slide is next in
order. The opening of the banking pins beyond that of drop
lock is called slide. Slide should be large enough to permit
freedom for escaping; usually }^ degree is considered suffi-
cient. Any amount more than this only increases the angular

114 A PRACTICAL COURSE IN HOROLOGY

motion of the lever and its connection with the balance,
resulting in an increased unlocking resistance, a shorter arc
of motion of the balance, and poor timekeeping. The banking
pins should be placed as far as possible away from the pallet
center so as to lessen the strain on the lever pivots should
the escapement overbank.

Slide is the last adjustment, the finishing touches, so to
speak, in escapement adjusting.

The Theoretically Correct Escapement

As sort of summary of our discussion of the escapement,
let us consider the specifications of a correctly designed
escapement.

When banked to the drop, the specifications should read
as follows :

Drop lock

V/2°

Safety lock

Va°

Corner freedom

Ya"

(1/°-

-Ya'-

= Ya")

Guard freedom

Ya'

(1/°-

-Ya"'

= Ya°)

With slide added, the specifications now read :
Total lock 2 ° (lJ^° + ^°=2°)

Safety lock ^°

Comer freedom \y^° (2° -

-^°-iM°)

Guard freedom 1^4° (2° -

-^°-iK°)

Slide lock ^° (2°-

-ly-y^)

Droplock 1>^° (2°-

-y^-iyn

Drawing the Lever Escapement

This chapter would not be complete without adding a few
lines about drawing the lever escapement. We have selected

ESCAPEMENT ADJUSTING

115

«>

c
bo

116 A PRACTICAL COURSE IN HOROLOGY

for this purpose the semitangental escapement shown in
Figure 40, as it comprises the best and latest design in
escapement construction.

The materials needed for drawing are : a drawing board,
drawing instruments, pencil, large and small protractors,
drawing paper, and India ink. A T square, two triangles,
and a French curve would be desirable although not absolutely-
necessary.

Study the drawing thoroughly before starting. Note at
what points the various angles originate. It is important,
also, to make the drawing on a large scale so as to minimize
the errors arising from imperfections in our drawing instru-
ments. Students who have no knowledge of mechanical
drawing would do well to read several chapters in any good
textbook on mechanical drawing.

Recommended texts:

Roberts, William E., Beginning Mechanical Drawing. The

Manual Arts Press, Peoria, Illinois, 1943.
Bennett, Charles A., Beginning Problems in Mechanical

Drawing. The Manual Arts Press, Peoria, Illinois, 1934.

Practical Application of Escapement Tests

Knowledge of the several escapement tests will be of little
value unless they reveal the necessary corrections to be made
in defective escapements. The purpose of this section is,
therefore, to show the application of the several tests as a
guide to escapement alterations. All of the examples listed
in the following pages are based on actual problems experi-
enced and corrected by the writer in the course of practical
work at the bench.

ESCAPEMENT ADJUSTING 117

In all cases it is first assumed that the escapement was
banked to the drop.

Adjustment 1.

Drop locks — correct
Guard freedom — excessive
Corner freedom — correct

In this escapement the guard freedom was greater than the
corner freedom. When the curve test was tried the roller
jewel would catch on the horns of the fork. The correction
consisted of flattening the end of the guard finger. For this
purpose a punch should be ground so that the end will fit
in the fork slot. A very light tap with a small hammer will
do the work. In some cases the guard pin may be lengthened
by forcing it further through the piece in which the pin is
placed. In this case, however, the pin could not be length-
ened, and, besides, the crescent in the safety roller was
rather wide. After flattening the guard finger the sides were-
stoned to provide the necessary guard freedom. The guard
safety test, the comer safety test, and the curve test were
tried and all tests were found satisfactory. The banking
pins were opened for slide.

Adjustment 2.

Drop lock — correct
Guard freedom — excessive
Corner freedmn — excessive

When the guard test and the comer test were tried the
pallets would recede from the tooth to such an extent that
the impulse faces of both tooth and pallets would show

118 A PRACTICAL COURSE IN HOROLOGY

contact; that is, the safety locks did not function on some
of the teeth. Since the drop locks were correct the excessive
guard and corner freedom suggested that the error was that
of a short fork. The lever was therefore stretched and the
guard test and corner test were again tried. It will be well
to state at this time that the stretching should be done a very-
little at a time, frequently making use of the tests. To
stretch a lever, place a small, flat- faced stump in the staking
tool and lay the lever thereon. Using a punch with sides
flattened, lightly tap the lever. If the lever becomes bent
during the act of stretching, turn it over on the stump and
give it another very light tap, using the same punch. Having
found the guard and corner tests satisfactory after stretching
the lever, the banking pins are opened for slide.

Adjustment 3.

Drop locks — deep

Guard freedom — satisfactory

Corner freedom — satisfactory

Since the drop locks were deep, the first act was to move
in both pallet stones and rebank to the drop. It was found
after rebanking to the drop that the roller jewel would not
pass in and out of the fork. The roller jewel was reset in a
position a little nearer to the balance staflf. Replacing the
balance, the corner test was tried and the comer freedom
was found to be correct, but when trying the guard test no
guard freedom was found. This example being an escape-
ment of the single roller type, the guard pin was bent away
from the roller table. The guard test, the comer test, and
the curve test were tried and all were found satisfactory, after
which the banking pins were opened for slide.

ESCAPEMENT ADJUSTING 119

Adjustment 4.

Drop locks — light
Guard freedom — none
Corner freedom — faulty

Since the drop locks were light the first act was to increase
the drop lock. This made it necessary to spread the banking
pins to a new banked-to-the-drop position. A proper guard
freedom and safety lock were found when trying the guard
test but when the corner test was tried the freedom was
found to be excessive and the locking was not safe on all
of the teeth. Examination showed that the roller jewel
tipped slightly toward the staff. The roller jewel was reset
in a position parallel with the staff and the corner test was
again tried, this time showing the correct comer freedom and
safety lock. As a final check-up, the guard, comer, and curve
tests were tried and all were found satisfactory, after which
slide was added.

Adjustment 5.

Drop locks — deep
Out of angle

When banked to the drop the escapement showed too much
guard freedom on the side of the receiving pallet and prac-
tically correct freedom on the side of the discharging pallet.
To correct the deep lock and to equalize the angular motion
of the lever from the line of centers, the discharging pallet
was moved in. The escapement was again banked to the drop
and the guard and corner tests were tried, showing too much
guard and corner freedom, also a complete absence of safety
lock. The drop locks were considered passable, so the error
was assumed to be that of a short fork. The lever w^as

120 A PRACTICAL COURSE IN HOROLOGY

stretched and the guard and corner tests were tried, showing
satisfactory guard and corner freedom and safety lock. The
banking pins were opened for slide.

Adjustment 6.

Drop locks — satisfactory

Out of angle

Drop and shake — close outside

The drop locks being practically correct, the first act was
to correct the condition of out of angle. As the out of angle
was slight with the access of freedom on the side of the
discharging pallet, the lever was bent in the direction toward
the receiving pallet. After banking to the drop and repeating
the several tests, the lever was found equal with regard to
the corner freedom, but the drop was still inequal, the
condition being that of close outside. It will be noted that
by bending the lever, the draw was increased, and for this
reason it was decided that the receiving pallet should be
tipped toward the pallet staff to equalize the drop. This was
accordingly done and a test for draw followed, showing a
satisfactory condition. All tests were tried and found
satisfactory. The banking pins were opened for slide.

Adjustment 7.

Drop locks — correct
Guard freedom — none
Corner freedom — none

After banking to the drop, this escapement did not show
any guard and corner freedom. Since the drop locks were
correct it was reasoned that the lever was too long. The
correction consists of grinding back the horns of the fork.

ESCAPEMENT ADJUSTING 121

This is done by fitting to the lathe an iron wire, part of
which is turned to fit the curve of the horns of the fork.
Using oilstone powder and oil, the horns are ground by-
holding the lever with a pair of cutting pliers against the
iron wire. Frequent applications of the corner test while
grinding prevented any possibility of overdoing the correction.
After grinding and polishing, the several tests showed a
very satisfactory escapement action. The banking pins were
opened for slide.

Adjustment 8.

Condition of escapement — satisfactory
Error — guard pin jams against roller table

In this example we haye single roller type of escapement.
After banking to the drop, the corner test was tried and
found satisfactory. When the guard test was tried, the guard
pin would jam or stick on the edge of the roller table. This
error, responsible for frequent stopping, occurs occasionally
in single roller escapements, but in double roller escapements
only when the guard finger is loose or bent. The correction
in the above example consisted of turning down and repolish-
ing the edge of the roller table and advancing the guard pin.
All tests showing satisfactory conditions, slide was added
and the watch proved to be an excellent timekeeper.

Adjustment 9.

Drop locks — deep

Out of angle

Draw — wanting on both pallets

This escapement was out of angle, with the excessive
guard freedom on the side of the receiving pallet, and since

122 A PRACTICAL COURSE IN HOROLOGY

the drop locks were deep and the draw was deficient it was
reasoned that all faulty conditions could be corrected by
moving in the discharging pallet. The discharging pallet
was accordingly moved in and after rebanking to the drop
a thorough examination showed that the drop locks were
correct; the lever's angular motion from the line of centers
was equal and the draw was satisfactory. This example
should be remembered. Errors of this kind occur quite
frequently and the correction is easy and the results are
certain.

Problems

i. What is meant by "banked to the drop" ?

2. Define drop lock.

3. State in degrees the correct amount of drop lock.

4. What is meant by "out of angle"?

5. How do you correct out of angle?

6. Define draw.

7. How do you examine draw?

8. Define drop.

9. How do you examine drop?

10. Define shake.

11. What is the lift?

12. What are the characteristics of a good action of lift?

13. What is meant by the expression "loss in the lift" ?

14. On what type of escapement is the loss in the lift greatest?

15. State in degrees the total angular motion of the lever when the
escapement is banked to the drop.

16. Should a watch run when banked to the drop ?

IT. In what manner do you go about making the guard safety test ?
corner safety test? curve test?

18. What is the purpose of the guard safety test? corner safety
test? curve test?

ESCAPEMENT ADJUSTING 123

19. What do the above tests aid in determining, if the drop locks
are correct ?

20. Define slide.

21. When is slide added?

22. Give the specifications for a correctly designed escapement, first
without slide, secondly with slide.

23. Does altering the banking pins change the drop lock ?

24. Define total lock.

25. Does altering the banking pins change the total lock ?

26. How do you correct an escapement in which the drop locks are
correct but there is an excessive guard and corner freedom ?

27. What is the error in an escapement that has a deep lock yet
the guard and corner freedom are satisfactory?

28. If the drop locks are light and the guard and corner tests show
no freedom, what is the correct procedure to put the escapement in
order ?

29. The drop locks are deep and the escapement is out of angle
with no guard and corner freedom on the side of the discharging
pallet. There is, however, too much guard and corner freedom on
the side of the receiving pallet. How would you go about correcting
this escapement ?

Part II
CHAPTER SEVEN

Cleaning and Oiling

Two METHODS are used in cleaning watches. The first
that we shall consider is the older method generally referred
to as the hand method. The second involves the use of the
cleaning machine.

The Hand Method

In using the hand method we proceed as follows : After
taking the movement apart string the larger pieces on a wire
loop, place the pieces in benzine, benzole, or any other good
cleaning preparation for several minutes. Remove and wash
pieces in hot water, using castile soap and a soft brush.
Rinse in clean water, dip in cyanide of potassium, rinse again
in clean water, immerse in alcohol, and dry in warm sawdust.
The same treatment is given to the wheels and other small
pieces that can be strung on the wire loop, but separately,
after the plates have been cleaned. The other small pieces
such as the pallet fork and the jewels, may be held against
a piece of hard pith or cork with a pair of tweezers especially
prepared for the purpose and brushed thoroughly with a
fine toothbrush previously dipped in benzole. After being
brushed, the pieces are dipped in alcohol and allowed to dry
on a sheet of watch paper. The balance must be cleaned
separately. The usual method is to dip the balance first in

124

CLEANING AND OILING 125

benzole, then in water followed by cyanide, again in water
and finally in alcohol, after which it is dried in sawdust.
The parts are now dry and the jewels should be rubbed with
two pieces of pegwood, one which has been pointed so as
to go through the holes and another that has been shaped
to fit the cups of the jewels.

The watch having been cleaned, the assembling and oiling
are next in order. Certain parts are oiled as the watch is
put together. For example, the main spring is oiled in the
barrel with clock oil. The winding mechanism, the escape
wheel, the pallet jewels, and the hole jewels where cap jewels
are used in connection are also oiled in the process of
assembling. The train is oiled after the watch is assembled,
also the center post that carries the cannon pinion. The
roller jewel is not oiled.

The W atch-cleaning Machine
The newer method of cleaning watches with the cleaning
machine has some advantages. It eliminates the use of
cyanide. Instead, an especially prepared cleaning solution
is used, together with water and a drying solution. The
machine is particularly satisfactory for the cleaning of small
pieces like the pallet fork, the jewels, and screws and has the
further advantage of eliminating the use of sawdust.

Briefly, the procedure is as follows : Place the parts in the
basket. There are provided several small spaces for the
small pieces and one large space for the plates. Lower the
basket into the jar containing the cleaning solution and allow
the motor to run for several minutes at a moderate speed.
Throw off the cleaning solution by raising the basket sufifi-
ciently to clear the solution. Lower the basket in a jar

126 A PRACTICAL COURSE IN HOROLOGY

containing water and rinse off the cleaning solution. Next,
lower the basket in the drying solution and run the motor as
before. Finally, allow the basket to spin in a receptacle
containing a lighted electric light bulb for a quick drying of
the watch parts. Thus the cleaning job is completed.

Problems

1. Is it important that the jewels should be cleaned with peg-
wood?

2. What parts of a watch do you oil?

3. What parts should not be oiled ?

4. Name some advantages in using the cleaning machine.

PART III
ADJUSTING

Part III
CHAPTER ONE

Preliminary Notes on Adjusting

There is a greater demand for watches of accurate time-
keeping qualities today than there was years ago. The rail-
roads require that their employees' watches run within certain
close limits and the complexity of modern life has shown a
need for greater accuracy also. To repair watches so that
close timing can be assured, a working knowledge of adjust-
ing is necessary.

The horologist who has never been concerned about the
theory and practice of adjusting has missed the real fascina-
tion and satisfaction of watch work. Aside from gaining
pleasure for himself, the repairman who applies the principles
of adjusting to his work will win the respect of his employer
and the sincere appreciation of his customers.

Adjusting consists in the execution of such manipulative
operations of the balance spring and other parts as to cause
a watch to function uniformly, the rate being within well-
defined limits under various conditions. Adjusting is nat-
urally divided into three branches: (1) position adjusting,
(2) isochronal adjusting, and (3) temperature adjusting.
These require independent methods of correction but in the
final analysis all three are inseparable when the work of
adjusting is completed.

128

PRELIMINARY NOTES ON ADJUSTING 129

General Observations

Before considering the more complex problems it will be
well to outline briefly those conditions which must be as
nearly perfect as possible before work can be attempted.

THE MAIN TRAIN

Close position and isochronal rating cannot be expected
unless the main train is in first-class condition. Extreme
variation is often caused by defects in any of the train wheels
and especially in the center wheel and mainspring barrel.
A correct amount of end shake and side shake is important.
It is well, therefore, to examine a number of high-grade
movements with regard to train freedom and note, also, the
recoil of the escape wheel after the train runs down when
winding the mainspring even to the slightest degree.

All train-wheel and balance pivots should be round and
well polished. A pivot that is not perfectly round will
function fairly well in a jewel hole that is round, but jewels
frequently do not have perfectly round holes. To show the
effect plainly, insert a three-cornered piece of steel in a
jewel that has a perfectly round hole. The triangular piece,
if it fits the jewel properly, will turn in the hole as perfectly
as a well-rounded pivot, but if we change the jewel for
one that is not round and repeat the experiment, the result
will be different. The triangular piece will become wedged
and will not turn. The effect exists in a lesser degree when
an imperfect jewel and an out-of -round pivot are used
together, yet the combined action of the two affects the time-
keeping qualities of a watch.

Furthermore, it is impossible to poise the balance if the
pivots of the staff are not round, and in this connection we

130 A PRACTICAL COURSE IN HOROLOGY

recommend the pivot polisher on all occasions where a
balance pivot needs reducing or polishing. This lathe attach-
ment forms the pivots perfectly round and cylindrical
(assuming the machine is properly adjusted), and it does
the work quickly and with a factorylike polish.

INFLUENCE OF THE LEVER ESCAPEMENT ON THE

ADJUSTMENTS

The impulse communicated to the balance through the
escapement should take place at the moment when the balance
spring is at its state of rest, that is, at that moment when
the spring is under no tension whatsoever. This ideal
condition would permit the balance and spring to perform
its arcs of vibration in the same time that a free balance
and spring would perform these arcs. However, the
mechanical means at one's disposal to keep the balance
vibrating does not meet the above requirements and one is
obliged to take account of the following laws.

1. An impulse delivered to a balance or pendulum before
the point of rest -will accelerate the vibrations.

2. An impulse delivered to a balance or pendulum after
its point of rest will retard the vibrations.

This principle can be easily demonstrated with a simple
pendulum. Impulse given to a pendulum before it reaches
its point of rest causes it to arrive at the point of rest more
quickly than if it were acted upon by gravity alone. Given
impulse after reaching the point of rest results in driving
the pendulum farther, resisting the force of gravity and at
no particularly accelerated rate, if any. Hence a retardation
takes place and the greater the distance the impulse takes
place after the point of rest, the greater is the retardation.

PRELIMINARY NOTES ON ADJUSTING 131

Now consider this factor in relation to the lever escape-
ment. The total angular motion of the lever is 10.5 degrees,
allowing for 2 degrees of lock. The relationship between
the acting length of the lever and the roller jewel radius is
3.5 to 1. The total lifting angle or contact with the balance
would be 3.5 X 10.5 or 36.75 degrees. Placing one half of
this figure on either side of the line of centers we would
have 18.375 degrees. However, the locking must be removed
from that portion of contact before the line of centers
(point of rest). Thus the impulse communicated before
the line of centers would be 18.375 —(3.5 X 2) = 11.375
degrees. The impulse after the line of centers would be the
full amount or 18.375 degrees.

According to the above analysis a retardation would result
for the short arcs. Further retardation occurs because of
the unlocking action, which is a serious resistance to the
free motion of the balance. Hence it is clearly seen that a
carefully adjusted escapement is of the utmost importance
in the fine position and isochronal rating of a watch.

WEIGHT OF THE ESCAPE WHEEL AND PALLET FORK
The escape wheel should be as light as possible consistent
with proper firmness. The sluggishness of a heavy escape
wheel directly increases the inequality of the impulse between
the receiving and discharging pallets owing to the inequality
of the draw and the lift on the pallets.

The lever, too, should be made as light as possible. It
was formerly the practice of manufacturers to add a counter-
poise to the lever, supposing that it was necessary to secure
close position rating. Research into the problem has demon-
strated that this is not necessary; in fact, it can be shown
that lightness of the lever is of much more importance. The

132 A PRACTICAL COURSE IN HOROLOGY

counterpoise only gives the force at the circumference of
the escape wheel more work to do and thus tends to make
the actions of the escapement and balance more sluggish.

MAINSPRINGS AND BALANCE MOTION
A mainspring unwinding in a barrel generally does not
take place in a concentric manner. Were it possible that
this ideal condition could be attained, there would be Httle
or no friction between the coils and a more smooth and
even motive power would result. This eccentricity in the
act of unwinding varies with the type of brace or hook used
on the outer end of the spring. Experience has shown that
any type of hook that maintains a quarter turn of the external
coil flat against the wall of the barrel gives good results and
it is gratifying to note that more and more manufacturers
are adopting some form of hook with this end in view.

It is important that the horologist use the very best main-
springs that money can buy. The superior performance of
a good spring is so apparent in the position rating of a watch
that no argument is necessary to convince the most skeptical.
Springs that are set, even to a comparatively slight degree,
should be replaced with new ones and the mainspring winder
should always be used.

The proper arc of motion of the balance is 540 degrees
when the watch is fully wound and lying in a horizontal
position. Horologists experience a feeling of real satisfaction
when a full balance arc is attained with the fitting of the
weakest possible mainspring. It is an indication that the
most even motive power will be maintained for the full 24
hours of running and that there will be only a slight falling-
off of the balance arc at the end of a 24 hours run.

PRELIMINARY NOTES ON ADJUSTING 133

How to ascertain the arc of motion. The question may
arise as to how we are able to ascertain this arc of motion.
This the eye can be trained to recognize easily and at a
glance by using the following method : Suppose the balance
is at rest with the roller jewel in the fork slot midway
between the banking pins. Now move the balance one half
of a circle or 180 degrees and stop. Release the balance
and the force of the spring will cause it to return to its point
of rest and 180 degrees farther on the opposite side. The
arc of motion would be 360 degrees. Again move the balance,
three fourths of a circle or 270 degrees, and allow it to
return on its own power to its point of rest and as far on
the opposite side. The arc of motion is now 540 degrees
and the balance will continue to vibrate between these points
as long as the proper motive power is maintained.

The arms of the balance become visible at the moment the
balance completes the arc of motion and starts in the opposite
direction on its return vibration. It is, therefore, at that
time when the balance stops that
the arc of motion can be deter- - o

mined.

With the assistance of Figure /3S^ | j^s

1, the problem can be more
clearly explained in this manner :

When the balance vibrates 180 /^{fm^m^imt^ ^a^^^Mi ^ **

degrees and returns to 0 degrees
and continues as far on the
opposite side — that is, when the
arms are visible at 180 degrees |

and 0 degrees — the arc of mo- 270"

tion is 360 degrees. When the Figure 1

134 A PRACTICAL COURSE IN HOROLOGY

balance vibrates 225 degrees and returns to 0 degrees and
continues as far on the opposite side — that is, when the arms
are visible first at 225 degrees and 45 degrees and on its
return vibration at 135 degrees and 315 degrees (forming a
cross at right angles) — the arc of motion is 450 degrees.
When the balance vibrates 270 degrees and returns to 0
degrees and continues as far on the opposite side — that
is, when the arms are visible at 270 degrees and 90 degrees —
the arc of motion is 540 degrees.

The horologist should examine the balance arc in all
positions and if the motion is faulty in certain positions the
necessary corrections should be made before any adjustment
to position or isochronism is attempted. Note carefully if
there is any difference between the arcs of dial up and dial
down. These positions should be equal. Note also the arcs
of pendant up, pendant right, pendant left, and pendant
down. The arcs for the vertical positions should be the
same although somewhat shorter than those for the two
horizontal positions, owing to the increased friction on the
balance pivots.

The arc of motion should never be longer than 540
degrees. Experience has shown that an arc longer than the
above figure produces a very fast rate for the first few hours
of running, after which time (the arcs becoming shorter)
the watch functions at its normal rate.

THE POISE OF THE BALANCE
One of the most common causes of variation between
positions is want of poise of the balance. The horizontal
positions are not affected but the error in the vertical positions
is considerable. The extent of the variation in the rate is
in proportion to the extent of the error in poise.

PRELIMINARY NOTES ON ADJUSTING 135

// the excess of weight is on the lower side of the balance
when at rest, the watch will lose when the arc of motion is
greater than 450 degrees, and will gain when the arc is less.

If the weight is on the top side when the balance is at
rest, the result will be reversed and the watch will gain when
the arc of motion is greater tlmn 450 degrees and will lose
when the arc is less.

Nature of error due to want of poise. Let us assume
that the excess of weight is on the lower side of the balance
when at rest. Suppose the balance vibrates at an arc of
almost 360 degrees, and in doing so the weight will stop near
the top of the balance. The force of the spring in returning
the balance to its point of rest will receive an added energy
in that of gravity acting on the weight. This means that
the spring will return to its point of rest a little more quickly
than when acted upon by the force of the spring alone. Now
assume that the weight, after having reached the bottom,
continues the arc on the opposite side. The force of gravity
acting on the weight is an added resistance to that of the
spring; in other words, the result of an added weight is,
in effect, the same as if a stronger spring were used and the
arc will be performed more quickly.

Now suppose that the motion is increased to 540 degrees
and in vibrating to this extent the weight starts from its
point of rest at the bottom and turns three quarters of a circle
and stops at right angles to a vertical line drawn through
the center of the balance. The force of the spring will
encounter a resistance due to gravity acting on the weight
as it starts upward toward the top, and after reaching the
top and starting downward, the force of gravity is an added
force downward. The effect would be a retardation during

136 A PRACTICAL COURSE IN HOROLOGY

the first portion of the path and an acceleration during the
latter portion to the extent that, for arcs above approximately
450 degrees the watch will lose. It would seem, therefore,
that at some point near to 450 degrees these forces would
counteract each other. Some authorities place this figure
at 444 degrees.

Poising the balance. Place the balance on a poising tool
and start the balance in motion. During the time it is rotating
hold a small compass as near as possible to the circumference
of the balance so as to ascertain whether or not it is mag-
netized. It is useless to attempt to poise a magnetized
balance; hence the first act is to demagnetize it if it shows
even the slightest bit of magnetism.

The balance having been demagnetized and found satis-
factory, proceed with the poising. Having located the posi-
tion of the heavy point it is good practice to take note of the
position of the regulator. This is our guide in deciding
whether we should reduce the weight at the heavy point or
add weight opposite to the heavy point, thus saving a certain
amount of timing after the watch is again running.

Should we find, when altering the weights, that the heavy
point has shifted a short distance we may feel certain that
good progress is being made. However, if it is found that
the heavy point has been shifted to the opposite side it is
evident that the correction has been overdone. It follows,
therefore, that in altering the weights we should proceed
with caution and thereby save much time, besides realizing
a much better piece of work.

Parallel pliers with cardboard glued to the jaws are very
effective for holding the balance while removing and replacing
the screws.

PRELIMINARY NOTES ON ADJUSTING 137

MAGNETISM

Magnetism is an ever-constant and insidious enemy to
horologists. The means by which a watch may be magnetized
are so numerous today that it is important that the repairman
form the habit of testing every watch for magnetism that
comes in for regulation, examination or repair.

In testing for magnetism place a small compass not only
over the balance but also over the winding wheels. The
mainspring being subject to magnetization as well as other
steel parts, has definite poles at the time the magnetic lines
of force passes through the watch. These poles are split
up into countless numbers as the mainspring unwinds. This
constant alteration of the relative position of the poles
between the mainspring, winding wheels, and the balance
helps to explain the erratic performance of magnetized
watches.

Theory of demagnetization. An alternating current is
that type of electrical current that changes its direction
constantly and when such current flows through a coil of
wire the poles also change. Figure 2 shows a conception of
an alternating current wave as it moves through time, and
the complete wave is called a cycle. An alternating current
of 60 such waves per second is said to have a frequency of
60 cycles.

When a steel rod is inserted in a demagnetizer and the
flow of alternating current is suddenly cut off, the steel rod
will be found to be magnetized, its poles being that of the
last half cycle that was sent through the wire. However,
if the current is left on and the rod gradually withdrawn,
the result will be different. It will be repolarized for every
cycle and each successive polarization will be weaker than

138 A PRACTICAL COURSE IN HOROLOGY

Figure 2. Wave form of alternating current

the preceding one. When withdrawn entirely from the field,
the magnetism has disappeared.

Demagnetizing a watch. The procedure in demagnetizing
a watch is much the same as explained above. Withdraw
the watch, keeping it central with the opening and giving it
a slight twist after leaving the opening. Continue the with-
drawal until the watch is about three feet or more from the
demagnetizer. If the first attempt fails to remove all mag-
netism repeat the operation.

Magnetized tools. The horologists' tools are subject to
being magnetized. Screw drivers and tweezers should be
closely watched. Avoid placing such tools in a north-south
position in or on the bench.

Problems

1. Why should balance pivots be made perfectly round ?

2. What is the effect of giving impulse to the balance or pendu-
lum before the point of rest? after the point of rest?

PRELIMINARY NOTES ON ADJUSTING 139

3. What is the effect of the lever escapement on the position
rating of a watch?

4. Is the counterpoise on the pallet fork of very great importance?

5. Does the type of fastening on the outer end of the mainspring
have anything to do with the friction between the coils?

6. What is the proper arc of motion of the balance ?

7. How do you ascertain the correct arc of motion?

8. What is the most common cause of position error ?

9. If the excess of weight is on the lower side of the balance
when at rest, will the watch gain or lose when running at an arc of
540 degrees ? when running at an arc of 350 degrees ?

10. What balance arc is considered the neutral arc?

11. How do you go about poising the balance?

12. How do you demagnetize a watch ?

Part III

CHAPTER TWO

Position Adjusting

The first portion of this chapter treats on position error
as related solely to the balance spring and of the effect of
gravity which is an ever-present force acting upon the inner-
most coils. The nature of this action is such that a positive
position error is produced.

A watch may be mechanically perfect — that is, its con-
struction from barrel to balance may be as exact as human
skill knows how to make it — and yet, in spite of such per-
fection, there will be a variation of from 15 to 30 seconds
in 24 hours between some two vertical positions due to the
condition of the balance spring alone. In watches that are
less perfect the error is frequently as high as 40 seconds or
more.

The Balance Spring and Its Poise Error

The oscillation of the inner portion of the spring cor-
responds very nearly to that of the collet ; that is, when the
collet travels three fourths of a circle, the first coil in the
center travels nearly an equal distance. It is further evident
that each of the several coils, as they tend to become more
distant from the center, will travel a shorter path until the
movement ceases altogether at the regulator pins. If the
coils are marked in a straight line from collet to regulator

140

POSITION ADJUSTING 141

pins, one would readily observe the distance traveled by the
several coils and the extent of their path under different arcs
of motion.

It is impossible to poise a spiral spring. Therefore it is
at once evident that it is the oscillation of the unpoised inner
portion of the spring, when acted upon by the pull of gravity,
that causes position error in the vertical positions. A state-
ment as to how this works need not be repeated here, for
the analysis given in the preceding chapter relative to the
poise error of the balance, applies to this condition also.
However, slightly varied effects under different arcs of
motion result, due to the fact that the greater mass of the
unpoised inner portion of the spring vibrates in a shorter arc
than does the balance proper.

Experimental demonstration. A demonstration at the
command of every horologist is to take several watches and
run them, first with the figure 1 up, and following with the
figures 2, 3, 4, etc., continuing the experiment around the
dial with all figures up, running the watches in each of the
12 positions for 24 hours and taking note of the rate in each
position. If an electric timing machine is available the
experiment can be made most conveniently and in a very
short time.

Table 1 shows the result of an experiment as stated above,
using four popular makes of American watches. All watches
were in excellent condition with balances perfectly poised,
fitted with theoretically corrected overcoils, and the grades
ranged from 17 jewels to 21 jewels. The arc of motion of
the balance of all watches was about 540 degrees when fully
wound and more than 450 degrees after 24 hours of running.

In watch Number 1 the rate was fastest at the time when

142 A PRACTICAL COURSE IN HOROLOGY

the figure 1 1 was up ; in watch Number 2 the rate was fastest
at the time when the figure 3 was up. The watches Number
3 and Number 4 had definite fast positions also, and in all
watches the slow position was opposite or nearly opposite that
of the fast position.

TABLE 1

NO. 1

NO. 2

NO.

NO. 4

Sees.

Sees

Sees.

+ 4

+ 3

+ 5

+ 8

+ 3

—

—11

— 7

—

— 8

+ 1

+ 5

— 3

+ 5

+ 4

.+ 6

+ 3

+ 8

THE NATURAL ERROR AND METHODS OF CORRECTION

An examination of the balance spring showed that the
fastest rate always occurred at the time when the middle of
the first half of the innermost coil happened to be up. This
error, due to the oscillation of the center of gravity of the
inner portion of the spring, is called the natural error and
is unavoidable. We can, however, make such alterations so
as to limit the fullest manifestation of the error in three
ways. These are as follows :

1. Pin the spring at the collet in such a position that the
natural error imll he the least detrimental to the uniform
rate of the watch.

POSITION ADJUSTING

143

Figure 3. Proper pinning at the collet. Figure 4. Proper pinning at the collet

2. Reduce the natural error by the application o, correct
terminal curves both outside and inside.

3. Neutralise the effect of the natural error by counter-
poising the balance.

These corrections will now be considered In the order
stated above.

The proper pinning at the collet. When fitting new
balance springs to pocket watches, certain pinning points
should be observed if the best position rates are to be
expected. The proper pinning at the collet is shown in
Figures 3 and 4. The first half of the innermost coil tends
upward as it leaves the collet in the direction of pendant up.

144 A PRACTICAL COURSE IN HOROLOGY

producing a fast pendant-up rate. It does not matter if the
spring tends to the left as shown in Figure 3 or tends to the
right as shown in Figure 4, for it can be readily seen that,
in either case, the middle of the first half of the innermost
coil stands in the direction of pendant up. When the spring
is pinned as stated above, the pendant-right and pendant-
left positions will have a slower but a nearly equal rate,
provided the balance spring is properly centered and vibrates
concentrically. The greater part of the natural error will
show up only in the pendant-down position, and since a pocket
watch in practical usage is seldom if ever subjected to this
position, it follows naturally that the pendant-down error is
of little importance.

Reducing the natural error. It was stated in the first
portion of this chapter that finely constructed watches vary
from 15 to 30 seconds in 24 hours between some two vertical
positions and watches that are less perfect would vary as
much as 40 seconds and more. If the natural error is shown
to be more than 30 seconds in 24 hours, the excessive
variation is due to want of perfection of the inner terminal
of the spring. A slight eccentric motion at the inner terminal
will cause a greater variation than would be the case if the
spring were perfectly true. Thus it is clear that the balance
spring should always be faultlessly trued at the collet and
equal attention should be given to both the flat and the round.

The Breguet type balance spring on position error. It
would now be natural for one to inquire as to the effect of
the Breguet spring with correct terminal as compared with
the ordinary flat spring on position error. Experiments
have demonstrated that the Breguet spring does reduce the

POSITION ADJUSTING 145

variation in the vertical positions, but only to a small degree,
proving that the position error is due primarily to the
oscillation of the center of gravity of the inner portion of the
spring.

Table 2 shows the results of an experiment using both
the flat and Breguet springs. The watch selected for this
example was an 18 size, 15- jewel grade, fitted with a flat
spring. The first column of the table shows the rate with
the flat spring, and the second column shows the rate with
the same spring after it was made over into a Breguet with
correct terminal. The watch was run in each position for
24 hours.

TABLE 2

FLAT

BREGUET

SPRING

Sees.

Pendant up

+ 7

Pendant right

+ 5

+ 3

Pendant left

— 6

Pendant down

—10

Counterpoising the balance. If the most perfect terminal
curves do not produce the desired results, counterpoising
may be tried. A general rule for the alteration is as follows :
Reduce the weight on the lower side of the balance in the
position that is slow. It is assumed that the balance has a
good motion and that at no time does the arc of motion fall
below 450 degrees during the 24 hours that the watch is
under observation. This is important if success in counter-
poising is to be expected. It should further be understood
that any alteration of the poise should be practiced only to
a limited extent ; otherwise, a most unsatisfactory and erratic

146 A PRACTICAL COURSE IN HOROLOGY

rate will result. Usually just a slight touch of the poising
saw will reduce the natural error as much as 5 to 10 seconds
in 24 hours.

Use of the Regulator Pins in Adjusting

The condition of the regulator pins play an important
part in the position rating of a watch. In fact, by slightly
opening or closing the pins as the case may require, it is
possible to bring the horizontal and vertical positions in
close agreement.

Let us suppose, for example, that the regulator pins are
opened slightly and the first coil of the spring vibrates equally
between the pins. We have literally made the active length
of the spring longer and the watch will go slower. It also
changes the rate between the long and short arcs. The effect
can be explained in this way :

Suppose that the balance is vibrating at an arc of 180
degrees and the first coil of the spring barely touches the
pins. For arcs below 180 degrees, the active length of the
spring will commence very nearly from the stud. Now, if
the arc of motion is increased to 540 degrees, the active
length of the spring will be shortened, commencing more
nearly from the pins. This will make the long arcs go faster
and the effect will vary in proportion to the changes taking
place in the arc of motion.

Suppose now that the pins are open, but instead of the
first coil of the spring vibrating equally between them, the
first coil leans against one of the pins. Assume that it
requires an arc of 360 degrees to lift the coil away from the
pin against which it leans. It is plain that for arcs below
360 degrees the active length of the spring will commence

POSITION ADJUSTING 147

from the pins, and for arcs above 360 degrees the active
length will commence more nearly from the stud. This
condition will make the long arcs go slower, or, in other
words, opposite to that in the former instance. Thus it can
be seen that the condition of the regulator pins may be the
cause of many of the disorders in the performance of watches.
It is also true that an intelligent manipulation of the pins is
the quickest and simplest means of correcting the variation
in the rate between the horizontal and vertical positions
because of the fact that the arc of motion is always shorter
in the vertical positions. The practical use of the regulator
pins is stated in the following rules :

1. If the regulator pins are closed and the watch gains
in the pendant-up position, a slower pendant-up rate is
obtained by opening the pins.

2. If the regulator pins are open and the watch loses in
the pendant-up position, a faster pendant-up rate is obtained
by closing the pins.

The spring should be so adjusted that the vibration of the
coil between the pins is equal ; otherwise, the coil will strike
one pin with more force than the other, and the result will
be very different from that stated in the above rules. Also
in spreading the pins, the vibration of the coil between them
should be very slight and discernible only with a powerful
glass. The pins should never be spread more than enough
to slow the mean rate 3 seconds an hour. If spread beyond
that amount the watch is apt to become a very unreliable
timepiece ; in other words, position adjusting by the manipu-
lation of the regulator pins can be practiced only to a limited
extent.

148 A PRACTICAL COURSE IN HOROLOGY

Adjustment to the Horizontal Positions

Effect of manipulating the regulator pins on the hori-
zontal positions. We have seen that the rates between dial
up and pendant up can be equalized by the manipulation of
the regulator pins. Should we alter the pins to secure the
desired results between dial up and pendant up, we could
expect a change in the rate between the horizontal positions
also. Often the horizontal rates are improved; sometimes
they are reversed. This would suggest that the adjustment
to the horizontal positions should be made last — that is, after
the corrections for dial up and pendant up are satisfactory.

After a little reflection it is evident that the difference
in the rate is due to the anisochronism of the balance spring.
The manipulation of the pins not only corrects the rate be-
tween dial up and pendant up, but more often than other-
wise it improves the isochronal rate also. This can be shown
by running a given watch in the position of dial up for 8
hours at an arc of 540 degrees and taking note of the rate
and then running it again for another 8 hours at an arc of
360 degrees. If the rate is slower when running at 360
degrees the pendant-up rate will usually be slower. Occasion-
ally there are exceptions.

Correction of errors in the horizontal positions. Errors
between the horizontal positions come generally under the
head of frictional errors and have to do with changes in
the arc of motion of the balance. A variation of 2, 3, or
even 4 seconds Is unimportant. Extreme variation can be
laid to rough pivots, dirt or thick oil, hole jewels that are
too small or too large, pitted cap jewels, balance pivots not
the same size, or a balance spring out of flat. As a rule the

POSITION ADJUSTING 149

fast position takes the shorter arc, which would suggest that
the position producing the fast rate is the one that calls for
correction. Assuming that the staff and jewels are as nearly
perfect as an inspection with a strong glass can determine, a
general rule for the correction of the rate in the horizontal
positions reads as follows : Round slightly the lower pivot in
the position that is fast.

Problems

1. Is it possible to poise perfectly a balance spring?

2. What is the natural error ? In what manner does it affect the
position rating of a watch ?

2. What is the proper pinning at the collet ?

4. Does the outer terminal of the balance spring have a greater
or smaller effect on the position rating of a watch as compared with
the inner terminal ?

5. Can the regulator pins be of use in adjusting a watch to posi-
tion? Explain.

6. What are the usual causes of position error in the horizontal
positions ?

Part III

CHAPTER THREE

Adjustment to Isochronism

The adjustment to isochronism is that adjustment which
has to do with the maintaining of a constant rate over a
definite period of time. Absolute isochronism is impossible
to attain because of several factors which are inherent in the
balance spring and for which there are no practical remedies.

A pendulum will make the long and short vibrations in
equal time. Start a pendulum in motion, traveling over a
given space in a given time. As the motion falls off, it will
be observed that the time consumed in each vibration does
not change. A slower or faster rate can be produced only
by lengthening or shortening the pendulum. The pendulum
is in reality a falling body and the laws which apply to
falling bodies apply to the pendulum also. Therefore adding
or reducing the weight of the pendulum does not affect the
time of vibration, for any change made in the mass carries
with it a proportional force in that of gravity.

Adding to or reducing the mass of the balance varies the
rate of vibration, for the strength of the balance spring does
not change. There are three factors upon which the time
of the vibration of the balance depends. They are :

1. The weight of the balance.

2. The diameter of the balance.

150

ADJUSTMENT TO ISOCHRONISM 151

3. The strength of the balance spring.

As already stated, the balance spring is the cause of the
more important disturbing factors in the isochronal rating
of watches. Of these disturbing factors, our attention will
first be directed to the problem of isochronism as affected by
varying the total length of the spring.

The Length of the Balance Spring on Isochronism

In every balance spring there is a certain length in which
the long and short vibrations are practically isochronal. Now
if this length is ascertained and we 'tnake the spring shorter
by whole coils, the short arcs will go faster; and if we make
the spring longer by whole coils, the short arcs will go slower.

It will be observed that the shortening or lengthening is
done only by whole coils. The reason for this is explained
in the following statement. If the spring is shortened by
some portion of a coil and not by whole coils, another effect
would take place which would alter the isochronism. For
example, the shortening of the spring by one half of a coil
may cause the short arcs to produce a losing rate instead of
a gaining rate. This problem is one that is distinctly separate
from the one now under discussion and it will be treated
more in detail in the next section.

A spring that is practically isochronal as far as the length
is concerned usually consists of about 13 coils. Springs
supplied by the manufacturer are correct for length. The
thought to remember from that which has been stated above
is that in fitting a new spring, the spring should not be
shortened excessively in timing, for in so doing the isochronal
rate is affected.

152 A PRACTICAL COURSE IN HOROLOGY

The Flat and Breguet Balance Spring
on Isochronism

The flat spring. If one will examine a flat spring in a
watch during the time the balance is in motion, it will be
observed that the vibration is wholly on one side and on the
side opposite the regulator pins. This is not all the eccentric
motion present in the spring, however, for a similar motion
takes place opposite the inner terminal, although it is less
visible to the eye. These eccentric motions affect the iso-
chronism because of several conditions. The principal ones
are: (1) a constant oscillation of the center of gravity,
(2) a persistent pushing and pulling at the balance pivots,
and (3) the effect of torsion, with which this section is
particularly concerned.

Action of the flat balance spring. Torsion is a circular
impulse which takes place at the innermost coils of the
spring. The result is a retardation or an acceleration, de-
pending on the relative positions of the two eccentric motions
as the balance vibrates. The anisochronism thus produced
can be varied by altering the length of the spring. Such
alterations, of course, change the angular distance between
the inner terminal and the regulator pins, and it is this change
of angular distance that decides the rate between the long and
the short arcs. The laws governing the isochronism as con-
cerned with the above statement are as follows :

1. When the angular distance between the inner terminal
and the regulator pins stands at even coils, that is, whole
coils, the short arcs gain.

2. When the angular distance between the inner terminal
and the regidator pins stands at even coils, plus half a coil,
the short arcs lose.

ADJUSTMENT TO ISOCHRONISM 153

Figure 5. Figure 6.

3. When the angular distance between the inner terminal
and the regulator pins stands at even coils plus one fourth or
three fourths of a coil, the long and short arcs are more
nearly isochronal.

Let us assume that the arc of motion of a given balance is
360 degrees, as an example of a short arc. If the spring is
pinned at even coils, the eccentric motions will stand in
opposite directions. According to rule 1, this produces a
gaining rate as compared with the long arcs. This can be
explained by reason of the fact that the eccentric motion of
the outermost coils exerts a force (when wound up) in the
direction of the arrow A, Figure 5, while the eccentric
motion of the innermost coil exerts a force in the direction of
the arrow B, and since these forces are in opposite directions,
there is a tendency toward acceleration as the arcs become
shorter than 540 degrees and the maximum is reached at 360
degrees. In unwinding, the forces are reversed but their
relation to each other is the same.

If the spring is pinned at even coils plus half a coil, the
eccentric motions will stand in the same direction, namely:
opposite the regulator pins. According to rule 2, this pro-
duces a losing rate as compared with the long arcs. Since the
forces of the eccentric motions are in the same direction,

154 A PRACTICAL COURSE IN HOROLOGY

Figure 6, there is less resistance or divergence of forces and
the balance may vibrate a little farther; hence a retardation
takes place.

If the spring is pinned at even coils plus one fourth or
three fourths of a coil, the eccentric motions will stand at
right angles to each other, the effects stated in rules 1 and 2
will be neutralized, and the watch will function at a more
nearly isochronal rate. Of course, when the watch is put to
practical usage, the motion of the balance is constantly
changing and this fact considerably complicates results.

The effect of torsion should not be confused with that of
the oscillation of the center of gravity, for the latter is a
function that is distinctly different and plays only a small
part in producing an anisochronism. The effect of torsion is
by far the most disturbing element and the only way its
effect may be reduced, aside from varying the terminal
pinnings, is the application of the most perfect terminal
curves, both outside and inside.

Reducing the isochronal error. The superior perform-
ance of the Breguet spring in the attainment of isochronism
is the reason for the passing of the flat spring. The above
analysis of the flat spring would at once suggest that it is
possible to vary the isochronism by manipulating the over-
coil of the Breguet spring so as to throw the eccentric motion
in some desired direction. That is correct reasoning; how-
ever, a spring that produces concentric vibrations will attain
close enough isochronism in most watches while at the
same time realizing the best position rating in the vertical
positions.

Modem watches are built with a better design and pro-
portion of parts than the older models and the correction of

ADJUSTMENT TO ISOCHRONISM 155

isochronism by means of altering a correct terminal is
seldom necessary. However, if the most perfect terminal
curves do not produce the desired results, the following rules
for altering the overcoil may be used.

1. If the short arcs are slow, bring in part of the body of
the spring and add it to the overcoil.

2. If the short arcs are fast, take part of the overcoil
and move it back into the body of the spring.

Problems

1. What is the effect of shortening a balance spring by whole
coils on the isochronal rating of a watch?

2. What is torsion ? What are the effects on a flat balance spring?

3. In what manner do you alter the overcoil to accelerate the
short arcs?

4. In what manner do you alter the overcoil to retard the short
arcs?

Part III

CHAPTER FOUR

The Adjustment to Temperature

In order to attain a practical system for the temperature
adjustment, the general practice has been to solder together
brass and steel for the rim of the balance. The brass occupies
the outer portion of the rim, consuming about three fifths
of the total thickness. The rim is cut near the arms to permit
the turning in and out of the loose ends, thus changing the
active diameter of the wheel. This movement compensates
for the changing elasticity of the balance spring during
temperature changes. This type of balance is known as the
compensating balance. A compensating balance made of
brass and Invar (a nickle-and-steel alloy) when used in
connection with a hardened steel balance spring has been and
is today the most satisfactory arrangement for combating
the temperature error.

However, there is a definite trend toward the use of a
plain uncut balance of a single metal and a balance spring
of a nonrusting, nonmagnetizing alloy, made principally of
iron, nickle, chromium, and tungsten, called Elinvar. A
slight temperature error exists, but there are certain advan-
tages that make it desirable to continue research for further
improvement. This is apparently the opinion of watch fac-
tories, for new models have appeared lately with definite

156

ADJUSTMENT TO TEMPERATURE 157

changes in the design of the movement as well as the balance
and spring, showing a persistent effort on the part of manu-
facturers to improve this new type of balance assembly.

Correcting the temperature error. Watches with com-
pensating balances are usually adjusted to temperature be-
tween 40 degrees Fahrenheit and 95 degrees Fahrenheit.
The rules for the adjustment are as follows :

1, If the watch runs slow in heat, move any even number
of screws that are opposite each other an equal distance
toward the loose ends of the rims.

2. If the watch runs fast in heat, move any even number
of screws that are opposite each other an equal distance
toward the balance arms.

Experience in effecting temperature adjustment is neces-
sary before the horologist is able to decide on the extent of
the alterations. More often than otherwise, several trials
are required. If the screws are moved considerable distance
the poise should be examined. Temperature adjustment is
practically permanent. The balance may be trued and poised
many times without interfering with the temperature adjust-
ment. However, in changing a balance spring, readjustment
would be necessary.

Equipment used for temperature adjusting. The equip-
ment used for temperature adjusting is an oven and a refrig-
erator. An oven suitable for the purpose need be nothing
more than a box fitted with an electric light bulb, a rheostat,
and a thermometer. The thermometer is placed inside and
in such a manner that it can be conveniently read through
an opening in the box. An ordinary electric refrigerator will
serve the purpose for the lower temperature.

158 A PRACTICAL COURSE IN HOROLOGY

Problems

1. Does the fitting of a new balance spring alter the temperature
adjustment of a watch?

2. A watch that is equipped with a compensating balance runs
slow in heat. How do you correct the error ?

3. Name the advantages of a solid, single-metal balance and
Elinvar balance spring.

Part III

CHAPTER FIVE

The Practical Work of Adjusting

The practical use of the instruction that has been written
in these pages will presently be demonstrated by several
examples taken from actual practice when the writer was
engaged in practical work at the bench. The data for the
watches to be adjusted are kept in a small notebook. The
date, customer's name, the make, size and grade of the watch
are recorded, followed by the several rates and an account
of the changes made to effect a satisfactory rating.

Practical Problems in the Adjustment to Positions

Preliminary notes. It will be necessary to briefly discuss
several factors before considering the concrete problems.
The method of computing the variation of the rate in the
different positions will be next in order and to assist in the
explanation, the reader is referred to Adjustment 4 on page
163. The first column, reading down, shows the progressive
rate — that is, the rate without setting the watch except at the
beginning of the test. The first figure shows a loss of 2
seconds, the rate for 24 hours, written — 2, the rate for dial
up. Without setting the watch, the figure below shows the
variation after running the watch in the position of dial
down, which is -\-2. The next figure below shows the varia-

159

160 A PRACTICAL COURSE IN HOROLOGY

tion after running in the position of pendant up, the rate
being — 19. For pendant right the progressive rate is — 30
and for pendant left the progressive rate is — 43.

The rate for each period of 24 hours is computed by sub-
tracting the lesser figure from the greater, provided the rates
are all plus or all minus. However, if the rates are plus and
minus, the figures are added and the sign before the last
rate added is prefixed to the 24-hour rate and entered in the
second column. To make this clear, we shall continue with
the example, explaining it in this way : The rate for dial up,
recorded as — 2, is carried to the second column. The rate
for dial down is computed by adding together — 2 and -j-2
and the rate is entered in the second column as -j-4, showing
the amount of the gain in excess of correct time. The rate
for pendant up is computed by adding together -\-2 and — 19,
and the result is entered in the second column as — 21,
showing the amount of loss deducted from correct time. The
rate for pendant right is the difference between — 19 and
— 30 and the rate for pendant left is the difference between
— 30 and — 43, both of which are shown in the second
column. Thus we may compute the 24-hour rate without
the necessity of setting the watch for every trial. Of course,
this is not necessary when the electric timing machine is used.

Maximum allowance in positions. Before testing in
positions the watch should be regulated to run within 15
seconds in 24 hours. The extent of variation between 5
positions in high-grade, 16-size watches should not exceed
6 seconds in 24 hours. For watches of a cheaper grade
and for most of the average grades in the 12 size, a variation
of not more than 10 seconds is considered passable. The
rate between the positions of dial up and pendant up are the

PRACTICAL WORK OF ADJUSTING 161

most important and should receive special consideration. The
rate between these two positions should not exceed 3 or 4
seconds ; in fact, it is not difficult to produce close agreement
and in most cases the manipulation of the regulator is all
that is necessary.

In the following pages are shown several examples of 3
and 5 position adjusting. These examples should be studied
carefully.

Adjustment 1.

Watch — 16 size, 23 jewels

Repairs — cleaned, staff fitted, balance poised.

After cleaning and repairing, the watch was tested in 3
positions and it was found to have a variation of 15 seconds
with a gain in the pendant-up position.

Dial up — 6 — 6

Dial down — 12 — 6

Pendant up — 3 +9 (15)

An examination showed that the spring was level and
properly centered and that the regulator pins were tightly
closed. Accordingly the pins were spread slightly and after
timing by turning in a pair of timing screws, the test was
again tried, this time with the following results.

Dial up — 2 — 2

Dial down — 5 — 3

Pendant up —9 — 4 (2)

Adjustment 2.

Watch — 16 size, 7 jewels

Repairs — Cleaned, main spring fitted

The first test showed a variation of 28 seconds as shown
below.

162 A PRACTICAL COURSE IN HOROLOGY

Dial up +8+8

Dial down +15 +7
Pendant up +48 +33 (28)

The error being considerable in the pendant-up position, it
was reasoned that the balance was out of poise. The spring
was removed and the balance was tested for poise. A slight
poise error was found, but the want of poise could not be
responsible for the whole variation of 28 seconds. There-
fore in addition to poising the balance, the regulator pins
were spread slightly and after timing the watch, the next
test showed a much improved rate.

Dial up +3 +3

Dial down +6 +3

Pendant up +5 — 1 (4)

Adjustment 3.

Watch — 16 size, 17 jewels
Repairs — cleaned, balance poised

This example showed a loss in the pendant-up position
instead of a gain as found in the previous examples.
Dial up — 2 — 2

Dial down — 4 — 2
Pendant up —18 —14 (12)

As the watch was in excellent condition throughout, the
error was most likely to be found in the regulator pins. An
examination showed that the pins were open and the correc-
tion consisted of closing the pins. A much improved rate is
shown on the second test.

Dial up —1 —1

Dial down — 2 — 1

Pendant up — 5 — 3 (2)

PRACTICAL WORK OF ADJUSTING 163

The adjustment to five positions consists in running the
watch in the positions of pendant right and pendant left in
addition to dial up, dial down, and pendant up. Because you
find a close rate between the three positions is no proof that
the watch will be an excellent timepiece. Errors could be
present that would show up only in the pendant-right and
pendant-left positions. Fine watches should be adjusted to
five positions, for often an unsuspected error in the general
construction of the watch is discovered.

Adjustment 4.

Watch — 18 size, 15 jewels
Repairs — cleaned

The first test showed the following rate.

— 2
+ 4
—21
—11
—13 (25)

The balance spring was pinned so that the fast position
stood in the direction of pendant up, yet the pendant-up rate
was slow. The dial-up and dial-down rates were also too
great to be allowed to pass.

The balance was placed in the truing calipers and the bal-
ance spring was examined and found to have an eccentric
motion at the collet. (Incidentally, the spring was removed
and the balance vv^as tested for poise, showing a slight poise
error with the excess of weight on the lower side of the
balance when in the pendant-up position.) The balance was
poised and the balance spring replaced and trued. Further
examination showed a slight vibration of the sprin.sr between

Dial up

Dial down

+ 2

Pendant up

Pendant right

Pendant left

—43

164 A PRACTICAL COURSE IN HOROLOGY

the regulator pins. Since the rates in the vertical positions
were all slow the pins were closed. The watch was regulated
to mean time and the results of the second test are shown as
follows :

—2

—1

—4

Dial up

Dial down

Pendant up

Pendant right

Pendant left

—5
-8 (7)

Adjustment 5.

Watch — 16 size, 21 jewels

Repairs — cleaned, staff fitted, balance poised

The horologist should be cautioned that there will be fre-
quently found watches that do not function satisfactorily in
position even though the balance and spring and the general
construction seem perfect in every detail. As already stated,
the watch must be mechanically in good order and it is
possible that some mechanical detail has been overlooked.
The example below was selected to show that the general
condition of the watch was responsible for the error in
position.

— 2

+ 8

+ 1
+ 10

—27 (37)

The first test, as shown above, was far from satisfactory.
This condition could not be due to want of adjustment of
the balance and spring as the balance was poised and the
spring was properly fitted, with the regulator pins closed.

Dial up

Dial down

+ 6

Pendant up

+ 7

Pendant right

+ 17

Pendant left

PRACTICAL WORK OF ADJUSTING 165

The arc of motion of the balance was examined and
found to be somewhat shorter than it should be. Since the
escapement was in excellent condition, it was assumed that
the main spring was set, and an examination showed that
such was the case. A new spring was fitted and the motion
was considerably improved. The train was examined and
a cracked jewel in the pallet bridge was discovered, al-
though the crack was so slight that it was discernible only
with a strong glass. A new jewel was fitted and without
doing anything else the next test showed a much improved
rate.

Dial up +2 +2

Dial down +3 +1

Pendant up — 1 — 4

Pendant right —1 0

Pendant left —7 —6 (8)

Adjustment 6.

Watch — 12 size, 17 jewels

Repairs — cleaned, balance poised, balance spring trued

This example shows that it is possible to have a close rate
between the horizontal positions and pendant up, yet the
pendant-right and pendant-left positions may be far from
satisfactory.

Dial up — 5 — 5

Dial down — 4 -|- 1

Pendant up — 3+1

Pendant right —23 —20

Pendant left —98 —75 (76)

The watch showed an abnormally slow rate in the pendant-
left position. It was observed that the arc of motion of the

166 A PRACTICAL COURSE IN HOROLOGY

balance was shorter than it should be. Further examination
showed that the drop locks were too deep, and, accordingly,
the drop locks were made lighter. The escapement was re-
banked to the drop and the necessary corrections made, yet
after the balance was replaced the arc of motion was still too
short. The mainspring was removed and found to be some-
what set and a little weaker than should be used in this grade
of watch. After replacing the mainspring with one of the
proper strength, the next test in position showed the follow-

ing results :

Dial up

—4

Dial down

Pendant up

Pendant right

—6

Pendant left

(7)

Adjustment 7.

Watch — 16 size, 23 jewels

Repairs — cleaned, staff fitted, balance poised.

The first test in positions showed the following rate :
Dial up 0 0

Dial down — 1 — 1

Pendant up + 4 -|- 5

Pendant right +17 +13
Pendant left +13 —4 (17)

The watch having a fairly satisfactory rate except for the
pendant-right position, it was reasoned that a slight counter-
poise would correct the error. Accordingly, the screw on the
lower side of the balance in the position that was slow, namely
pendant left, was reduced very slightly. The next test showed
the following results :

Dial up

Dial down

Pendant up

Pendant right

Pendant left

PRACTICAL WORK OF ADJUSTING 167

—2
+2
+3
+2 (5)

It will be observed that the counterpoise retarded the fast
position more than it accelerated the slow position. This is
always the case. The excess of weight, when placed below
the center of gravity, will retard the rate more than the same
weight, when placed above, will accelerate the rate. This is
a point to remember when attempting the correction of a
position error by counterpoising.

Adjustment 8.

Watch — 12 size, 17 jewels

Repairs — cleaned, balance poised, new balance spring
fitted

In this example the balance spring was selected from a
stock of uncolleted springs. The spring was vibrated, coUeted
and trued at the collet, overcoil formed, and corrected to pro-
duce concentric vibrations. The first test showed the follow-
ing rate.

+ 5
—10
+17
+ 10
+ 12
—15 (32)

The natural error being considerable. It was decided that
a new inner terminal should be made. About three inner
coils were broken out and the Lossier inner terminal was

Dial up

+ 5

Dial down

Pendant up

+ 12

Pendant right

+22

Pendant left

+34

Pendant down

+ 19

168 A PRACTICAL COURSE IN HOROLOGY

formed. Noting also that the dial-up position was fast the
lower pivot was rounded slightly. After timing the watch,
the second test showed the following rate.

Dial up 0 0 Pendant right — 5 — 5

Dial down 0 0 Pendant left — 8 — 3
Pendant up 0 0 Pendant down —18 —10 (10)

This example with the Lossier inner curve is given to show
how the natural error can be reduced. The very best rates
can be attained only with poised collets and theoretical inner
terminals. The above watch when carried by the owner pef-
formed at a rate within ten seconds a month.

Practical Problem in Isochronal Adjusting

In the chronometer it is possible to attain isochronism by
altering the form of the terminal curves or by selecting a
certain relation of the pinnings between collet and stud. How-
ever, any attempt made to effect isochronism by these meth-
ods in watches may seriously interfere with the position
adjustment. As the position adjustment is more important,
it is desirable to sacrifice the isochronal rating when both
cannot be satisfied. If the spring is pinned correctly for
position adjustment, the best pinning for isochronism may
or may not exist, depending on the length of the spring and
the design of the watch. As stated in Chapter Three,
Isochronal Adjusting, the best we can do to attain practical
isochronism in watches lies in the correct formation of the
terminal curves.

Balance springs supplied by the manufacturer. When
fitting a spring supplied by the manufacturer for a certain
model of watch the proper length need not be considered.

PRACTICAL WORK OF ADJUSTING 169

for that factor has already been taken care of by the maker.
There are occasions, however, when the spring in the watch
has been considerably shortened by some workman who was
not acquainted with the laws of isochronism as governed by
the length of the spring. In such cases the spring must be
replaced with a new one of the proper length, if practical
isochronism is to be expected.

Method used in testing isochronism. After being wound
and set, the watch is run for 6 hours, after which time the
rate is recorded. The watch is then run for 24 hours from
the time it was wound and set, when rate is again recorded.
The watch is run 6 hours longer without winding and the
rate is recorded for the third time. The rate for the first
and last period of 6 hours is separately computed for a
period of 24 hours. In this manner the variation of the rate is
shown for the long and short arcs.

A practical problem. In showing an example of isochronal
adjusting it is possible only to prove that which has already
been stated. The following example, therefore, shows how
the correction of the eccentric motion of the balance spring
improves the isochronism. The spring had an eccentric motion
in the direction opposite the regulator pins and the first test
showed the following results :

[2 :00 P.M. set 0

8:00 P.M. —1

Rate in 24 hours — 4
'2 :00 P.M. next day —3
8 :00 P.M. —7

Rate in 24 hours —16 (12)

The first test showed a loss of 12 seconds in the short arcs.

Long arcs

Short arcs

170 A PRACTICAL COURSE IN HOROLOGY

After correcting the overcoil so that the vibrations of the
spring were concentric to the center of the balance, the next
test showed a much improved rate :

'9 :(X) A.M. set 0

3:00 P.M. 0

Rate in 24 hours 0

'9 :00 A.M. next day +1
3 :00 P.M. +1^

Rate in 24 hours —2 (2)

Long arcs

Short arcs

Practical Problem in Temperature Adjusting

Of all the adjustments of watches temperature adjustment
is the best understood, no doubt because of the fact that it
has been the principal error to be eradicated in the chronom-
eter and therefore studied more than the other.

One example of temperature adjusting will suffice, as the
correction in all cases consists merely of moving opposite
pairs of balance screws from one position to another.

The example shown below had a rate of 20 seconds fast in
heat:

Cold— 5 Heat +15

To show the location of the screws and the position to
which they are moved, it is customary to number the screw
holes. The holes nearest the arms are numbered 1, the next 2,
the next 3, etc., the highest number indicating the last holes
near the loose ends of the rims.

Since this example showed a gain in heat, the correction
consisted of moving the screws at the free ends of the rims
toward the balance arm. Accordingly, the screws in holes
number 9 were moved to holes number 5, and the screws in

PRACTICAL WORK OF ADJUSTING 171

holes number 11 were moved to holes number 9. The next
test showed a variation of 10 seconds slow in heat.

Cold +10 Heat 0

This shows that the correction was overdone. Therefore
the screws in holes number 9 were moved to holes number 11.
The next test as recorded below shows a very satisfactory
rate.

Cold +6 Heat +5

Final Timing and Regulating

Timing screws and washers. After a watch has been
cleaned and repaired, the variation in time should not be
corrected by moving the regulator, but rather by turning in
or out the timing screws as the watch may require. Some
watches do not have timing screws and the correction must
necessarily consist in undercutting the balance screws or
adding timing washers. H the watch runs within 30 seconds
in 24 hours the regulator may be used.

The middle temperature error. It was stated in Chapter
Four of Part I that the middle temperature error causes a
variation of from 2 to 6 seconds, being faster than the rates
between the extremes of heat and cold. It is better, therefore,
to regulate the watches on the rack to run a few seconds fast
rather than just on time, for the watches on the rack are
running in the normal temperature and the middle tempera-
ture errort is in effect. When carried in the pocket the
higher temperature would cause the watch to run slow.

Delivering the watch. When delivering a watch to the
customer tell him that it is preferable to wind the watch in

172 A PRACTICAL COURSE IN HOROLOGY

the morning. There is a good reason for so doing. The best
part of the mainspring is used during the day when the
watch is carried. Since the balance has a sHghtly shorter arc
of motion when running in a vertical position, it is better
to take advantage of the extra power that would be avalable
by winding in the morning. Because a watch is usually laid
flat on a table or dresser at night it is at once evident that a
more nearly uniform balance arc takes place if the above
practice of winding the watch is adhered to.

The horologist should take time to explain to the customer
that there will probably be a variation of several seconds
during the first few weeks of carrying the watch and ask the
party to come in, in a week or two, for comparison with the
correct time and for further regulation if necessary. Tell
the customer that it takes several weeks to properly regulate
a watch and that he may come in as often as he finds it con-
venient. In this manner much of the dissatisfaction of the
repair department is eliminated besides making many friends
for the store.

GLOSSARY OF TERMS

Addendum. The portion of a tooth of a wheel or pinion
beyond the pitch circle.

Arbor. Axis of the balance wheel or mainspring barrel.

Balance. The vibrating wheel of a watch, which, in con-
junction with the balance spring, regulates the progress of
the hands.

Balance arc. A part of the vibration of a balance.

Balance cock. The support for the upper pivot of the bal-
ance staff.

Balance spring. A long fine spring that regulates the vibra-
tion of the balance.

Balance staff. The axis of the balance.

Banking pins. Two pins that limit the angular motion of
the lever.

Banking to the drop. An adjustment of the banking pins
permitting the escape wheel teeth to drop off the pallets.

Barrel. A circular box for the reception of the mainspring.

Barrel arbor. The axis of the barrel, round which the main-
spring coils.
Beat. One vibration of the balance and spriiig.

Beryllium alloy. An alloy of iron, nickel, and a small
percentage of beryllium, used for balance and balance
spring.

173

174 A PRACTICAL COURSE IN HOROLOGY

Breguet spring. A balance spring in which the outer coil is
raised above and carried over the body of the spring.

Bridge. Upper plates in a watch for the support of the
wheels.

Butting. Two wheels touching on the points of the teeth
when entering into action with each other.

Cannon pinion. The pinion with a long pipe to which the
minute hand is fixed.

Center of gravity. That point in a body around which the
mass is evenly balanced.

Center wheel. The wheel in a watch the axis of which
usually carries the minute hand.

Chronograph. A watch that has a center-seconds hand
driven from the fourth wheel which can be started, stopped,
and caused to fly back to zero by pressing on a knob or
lever.

Chronometer. A boxed timepiece for use on ships at sea.

Chronometer escapement. A spring detent escapement used
in chronometers. Invented by Pierre LeRoy about 1765.

Circular escapement. An escapement so constructed that
the central portion of each pallet stone's impulse face
stands at an equal distance from the pallet center.

Circular pitch. The pitch circle divided into as many spaces
as there are teeth on the wheel or pinion.

Club-tooth wheel. That type of wheel which has a lifting
face off the end of the teeth.

Clutch pinion. The pinion surrounding the square of the
stem. Serves alternately to wind and set the watch.

GLOSSARY OF TERMS 175

Collet. A small, circular piece fitting friction-tight to the
balance staff and which is pierced to receive the inner coil
of the balance spring.

Compensating balance. A balance the rim of which is made
of brass and steel. The diameter is caused to increase or
decrease in different temperatures, so as to compensate for
changes in temperature on both balance and spring.

Corner safety test. A test to show the presence or absence
of safety lock when the slot corner is brought in contact
with the roller jewel.

Crescent. A circular notch in the edge of the roller table

for the reception of the guard pin or finger.
Crown. A grooved circular piece fastened to the stem for

winding the watch.

Crown wheel. A wheel that drives the ratchet wheel.

Curb pins. Regulator pins.

Curve test. A test used to discover if the curves of the
lever horns are correctly related to the roller jewel.

Cylinder escapement. A frictional escapement patented by
Thomas Tompion 1695.

Dedendum. The portion of the tooth of either wheel or
pinion inside of the pitch circle.

Dial train. A train of two wheels and two pinions that con-
trol the progress of the minute hand and the hour hand.

Diametrical pitch. The diameter of the pitch circle divided
into as many spaces as there are teeth on a wheel or pinion.

Discharging pallet. The pallet stone which an escape-
wheel tooth slides over in order to emerge from between
the pallet stones.

176 A PRACTICAL COURSE IN HOROLOGY

Distance of centers. The distance on a straight line from
center to center, as between balance center and pallet center.

Double-roller escapement. A form of lever escapement in
which a separate roller is used for the safety action.

Draw. A force that keeps the lever against the banking pins,
created by the slant of the pallet stones.

Driven. The mobile that is being forced along by the driver.

Driver. The mobile that forces the other along.

Drop. The free motion of the escape wheel after impulse to
the pallets has been given.

Drop lock. The extent of the lock on the pallets after an
escapement has been banked to the drop.

Duplex escapement. A watch escapement in which the
escape wheel has two sets of teeth. One set locks the wheel
by pressing on the balance staff. The other set gives im-
pulse to the balance. The balance receives impulse at every
other vibration. Accredited to Pierre LeRoy about 1750.

Epicycloid. A curve generated by a point in the circum-
ference of a circle as it rolls upon another circle. It forms
the kind of tooth used in watch wheels.

Equidistant escapement. An escapement so constructed
that the locking faces of the pallet stones stand at an equal
distance from the pallet center.

Elinvar. A nonrusting, nonmagnetizing alloy containing
iron, nickel, chromium, tungsten, silicon and carbon. Used
for balance and balance spring.

Fork. The horns and slot of the lever.

GLOSSARY OF TERMS 177

Fourth wheel. The wheel of a watch that drives the escape
pinion.

Guard pin or finger. A pin or finger working in and out of
the crescent to preserve the safety action.

Guard safety test. A test to show the presence or absence
of safety lock when the guard pin or finger is brought in
contact with the edge of roller.

Heel of tooth. Letting-oif corner of a tooth of the escape
wheel.

Horns. The circular sides of the fork leading into the slot.

Hour wheel. The wheel that carries the hour hand.

H3rpocycloid. A curve generated by a point in the circum-
ference of a circle when it is rolled within another circle.

Impulse pin. Roller jewel.

Invar. A steel alloy containing about 36 per cent nickel.
Used in the making of balance wheels.

Isochronism. The property of a balance spring that allows
it to perform the long and short arcs in equal time.

Letting-off comer. Corner of a pallet stone from which a
tooth lets off.

Lever. A metal piece attached to the pallets that carries
impulse to the balance.

Lever escapement. A watch escapement that delivers im-
pulse to the balance by means of two pallet stones and a
lever. The extremity of the lever has a forked slot that
acts directly on a roller jewel which is attached to the
balance. Invented about 1750 by Thomas Mudge.

Lift. The pitch or slant of a tooth or pallet stone.

178 A PRACTICAL COURSE IN HOROLOGY

Line of centers. A line drawn from center to center, as of
any wheel or pinion.

Locking. The overlapping of a tooth on a pallet stone.

Lossier curves. The theoretical outer and inner terminals
as designed by L, Lossier.

Main train. The toothed wheels in a watch that connect the
barrel with the escapement.

Middle-temperature error. The temperature error between
the extremes of heat and cold characteristic of a com-
pensating balance and steel balance spring.

Minute v^^heel. The wheel driven by the cannon pinion.

Out of angle. Unequal angular motion of the lever from the
line of centers when an escapement is banked to the drop.

Overbanked. A term used to indicate that the lever escape-
ment is out of action.

Overcoil. The last coil of the Breguet spring that is bent
up and over the body of the spring.

Pallet arms. The metal body which contains the pallet stones.

Pallets. The metal body attached to or a part of the lever.
The term includes the pallet arms and pallet stones.

Pallet staff. The axis of the pallets.

Pallet stones. Jewels or stones inserted in the pallet arms.

Phillips* spring. A balance spring with terminal curves
formed on lines laid down by M. Phillips. The term
"Phillips' curve" is rarely used.

Pinion. The smaller wheel with teeth called leaves, working
in connection with a larger wheel.

GLOSSARY OF TERMS 179

Pitch circle. A circle concentric with the circumference of
a toothed wheel and cutting its teeth at such a distance
from their points as to touch the corresponding circle of
a pinion working with it and having with that circle a
common velocity, as in a rolling contact.

Pitch diameter. The diameter of the pitch circle.

Pivot. The end of a rotating arbor.

Plate. Discs of brass or nickel which form the foundation
of a movement. The lower plate lies next to the dial. The
upper pieces supporting one, two, or three wheels are
generally referred to as bridges. In the full-plate watch
the upper piece is called the top plate.

Potence. A hang-down bracket used for supporting the
lower pivot of the balance staff in full-plate watches.

Quarter screws. Four screws used in timing.

Ratchet wheel. A wheel that is fastened to the barrel arbor.

Ratchet tooth wheel. The name given to the English type
escape wheel which has pointed teeth.

Receiving pallet. The pallet stone over which a tooth of
the escape wheel slides in order to enter between the pallet
stones.

Remaining lock. The lock remaining when the guard and
corner tests are tried. More often called "safety lock."

Repeater. A watch that strikes, having two hammers and
two gongs. A lever Is provided to set the striking mech-
anism into action. A quarter repeater strikes the hour and
the last quarter hour. A minute repeater, in addition,
strikes the number of minutes since the last quarter.

180 A PRACTICAL COURSE IN HOROLOGY

Right-angled escapement. An escapement in which the Hne
of centers of the escape wheel and pallets are at right angles
to pallets and balance.

Roller jewel. A long, thin jewel inserted in the roller table ;
sometimes called impulse pin.

Roller table. A circular disc attached to the balance staff
in which is fitted the roller jewel.

Run of lever. The motion of the lever toward the banking
pins when slide is present. Run always equals slide.

Safety lock. The lock remaining when the guard and corner
tests are tried.

Semitangental escapement. An escapement where the lock-
ing face of the receiving pallet is planted 31 degrees from
the line of centers and the discharging pallet 29 degrees
from the line of centers. The receiving pallet locks only on
the tangent.

Shake. The space separating the letting-off comer of the
pallet from the heel of a tooth when the opposite pallet is
locked at the lowest locking corner. Shake is always less
than drop.

Single roller escapement. A form of lever escapement in
which one roller performs the functions of both impulse
and safety actions.

Slide. The opening of the banking pins beyond that of drop
lock.

Steady pins. Pins used to secure the perfect alignment of
two pieces of metal.

Stem. The winding arbor of a watch.

GLOSSARY OF TERMS 181

Stop work. A mechanical device for preventing the over-
winding of a mainspring.

Straight line escapement. An escapement in which the
centers of the escape wheel, pallets, and balance are planted
in a straight line.

Stud. A small piece of metal pierced to receive the outer coil
of the balance spring.

Third wheel. The wheel of a watch that drives the fourth
pinion.

Timing screws. Screws used to bring a watch to time,
sometimes called the mean-time screws.

Toe of tooth. Locking corner of a tooth of the escape wheel.

Total lock. Drop lock with slide added.

Train. A combination of two or more wheels and pinions,
geared together and transmitting power from one part of
a mechanism to another.

Tripping. A tooth of the escape wheel running past the
locking face of a pallet stone at a time when safety lock
should be present.

Wheel. Any circular piece of metal on the periphery of
which teeth may be cut of various forms and numbers.

Winding pinions. A pinion surrounding the stem that
drives the crown wheel.

WATCHES

Borer and Bowman,

Modern Watch Repairing and Adjusting

DeCarle, Donald,

With the Watchmaker at the Bench

Garrard, F. J.,

Watch Repairing, Cleaning and Ad j testing

Gribi, Theo.,
Practical Course in Adjusting

Grossman, Jules and Herman,
Lessons in Horology

Hood, Grant,

Modern Methods in Horology

Kleinlein, Walter J.,

Rules and Practice in Adjusting Watches
Practical Balance and Hairspring Work

Thisell, A. G.,

Watch Repairing Simplified

Wilkinson, T. J.,

The Escapement a7id Train of American Watches

182

BIBLIOGRAPHY 183

WATCHES AND CLOCKS
Britten, F. J.,

Watch and Clock Makers* Handbook

Haswell, Eric,
Horology

Saunier, Claudius,

Treatise on Modem Horology

CLOCKS
Garrard, F. J.,

Clock Making and Repairing

Gordon, G. F. C.,

Clock Making, Past and Present

Langman and Ball,
Electrical Horology

Philpott, Stuart F.,
Modern Electric Clocks

Robinson, T. R.,

Modern Clocks, Their Repair and Adjustment

HISTORY
Britten, F. J.,

Old Watches and Clocks and Their Makers

Chamberlin, Paul M.,
If s About Time

Gould, Rupert T.,

The Marine Chronometer, Its History and Developinent

184 A PRACTICAL COURSE IN HOROLOGY

Hering, D. W.,

The Lure of the Clock

Nutting, Wallace,
The Clock Book

ALLIED SUBJECTS

Bennett, Charles A.,

Beginning Problems in Mechanical Drawing

Eaton and Free,

Machine Shop Science and Mathematics

Feirer and Williams,
Basic Electricity

McMackin and Shaver,
Mathematics of the Shop

Roberts, William E.,

Beginning Mechanical Drawing

INDEX

Action PAGE

fork and roller 47, 110

four-to-one roller 47

of flat spring 152

three-to-one roller 47

safety in 45, 110

unlocking and impulse. 47

pallet on wheel 44, 100

wheel and pallet_44, 100, 106

Addenda 33

Adjusting 128-172

escapement 100-113

isochronism

128, 150-155, 168, 169

position 128, 140-149

practical work of 159

regulator pins 146, 147

temperature

128, 156-158, 170

Altering drop locks 102

Angle, escapement, out of

102,119,120,121
Angular motion of lever_ liO
Arc of motion of balance. 133

185

PAGE

Balance 14, 52-53

axis 84

counterpoising 145

motion 132

pivots 129

poise of 134, 136

Balance spring 14, 53, 58

action of flat 152

and its poise error 140

Breguet type 144

eccentric motion of 97

fitting of 88, 89

flat 58

isochronal condition of _ 150
length of, on ischronism 151

on position error 144

truing of 97

Balance staff 73-84

Banking pins 41-101

Banking to drop 101

Barrel 14,24

Beat 14

Bell-metal laps 7S

186

A PRACTICAL COURSE IN HOROLOGY

PAGE

Bezel-type jeweling 68, 70

Boxwood laps 79

Breguet balance spring

58,90,144,152

Burnisher 69

Butting 63

Butting error, example of 121

Calculating

a new train 23

beats 18,88

length of mainspring 26

length of overcoil 91

number of hours watch

will run 24

teeth and leaves of miss-
ing mobiles 20-22, 30

thickness of mainspring 25

turns of pinion 15

turns of train 16-18

Cannon pinion 28

Cast-iron laps 78

Cement chucks 82, 85

Center of gravity 59-141

Center wheel and pinion. 16
Centers

distance of 32

line of 33

Circular pallets 42

PAGE

Circular pitch 33, 37

Cleaning watches 124-126

Close inside 104

Close outside 104, 120

Club-tooth escapement

45, 107, 108
Qub-tooth escape wheeL 39

Collet 77, 80, 84, 143

Comments on fast trains- 19, 20

Compensating balance 53

Cone pivot 80

Controlling mechanism _52-60
Converting flat spring into

Breguet 94,95

Corner freedom 117, 120

Comer safety test 110, 112

Counterpoising 145

Crescent 41, 50

Curve test 113

Dedenda, dedendum

33, 34, 175

Degrees of angular motion 1 10
Degrees of lift 110

Delivering the watch 171

Demagnetizing a watch__ 138

Depthing too deep 62, 64

Depthing tool 35

Depthing too shallow 62, 64

INDEX

187

PAGE

Dial train 15, 28-31

Diameter, full 32, 36

pitch 32

Diametrical pitch 33

Dimensioning, three and

five position 161-164

Discharging pallet 14,40

Distance of centers 32

Double roller 41, 50

Draw 45, 103, 104

Drawing lever escapement 1 14

Driven 33

Driver 33

Drop 46, 104

banking to the 101

close outside 120

Drop lock 101, 102

deep 118,119,121

light 119

Eccentric motion of bal-
ance spring 97

Effect of manipulating

regulator pins 146

Elinvar 57, 156

Epicycloid 33

Equidistant pallets 41

Equipment for tempera-
ture ad j usting 157

PAGE

Errors common in staff

making 83

Escapement 14, 39-51, 100-123

adjusting 100-123

double-roller 50

drawing 114

lever 44

semitangental 43

single-roller 49

tests 116

theoretically correct 114

Escape wheel 16, 39

number of teeth of 43

weight of 131

Fast pendant-up 161

Fast trains 19

Final timing and regulat-
ing 171

Finishing end of balance

pivots 83

Fitting balance springs 88-99

Five positions, adjustment

to 163

Flat spring

on isochronism 152

on position error 144

Flat springs, fitting 89

Fork and roller action 47-1 10

188

A PRACTICAL COURSE IN HOROLOGY

PAGE

Fork slot 40, 47

Four-to-one roller action. 47
Fourth wheel and pinion_ 16
Friction jeweling 70, 71

Gearing 32-38, 62

Gravers

for setting jewels 68, 69

for turning staff 73, 74

Grinding and polishing

cone pivot 80

Grinding materials 79

Guard freedom

excessive 117

none 119

Guard pin jams 121

Guard safety test 110

Guard tripping error 117-119

Hand method of cleaning

watches 124

Heel of tooth 40

Horizontal positions, ad-
justment to 148

Horns of fork 40

Hours of running 24

Hour wheel 28

PAGE

Hub, turning and rough-
ing 81

Hypocycloid 34

Impulse 130

Impulse and unlocking ac-
tion 47

Impulse face 40

Influence of the escape-
ment on the adjust-
ments 130

Invar 57, 156

Isochronal adjustment,

example of 168

Isochronal error 154

Isochronism

adjustment to 150-155

length of balance spring

on 151

testing to 169

Jeweling : 68-72

Jewel pin or roller jeweL40, 41

Laps 78,79

Length of balance spring- 151
Letting-of¥ corner 40

INDEX

189

PAGE

Lever 40

angular motion of 110

escapement

39-51,114,116,130

Lift 45,105,107,110

Line of centers 33

Lock

drop safety, tests

14,101,110-113

total 114

Locking 44

Locking face 40

Lossier inner and outer

terminals 59, 60, 91, 95

Machine method of clean-
ing watches 125

Magnetized tools 138

Magnetism 137

Main train 15, 129

Mainspring

length of 26

space in barrel 26, 27

thickness of 25

Mainsprings 24, 132

Maximum allowance in

positions 160

Measuring for balance

staff 75

PAGE

Middle-temperature error

54, 171

Minute pinion 28

Minute wheel 28

Motion of balance 133

Motion of balance spring. 97

Natural error 142, 144

Narrow roller jewel 49

Number of teeth of escape

wheel 43

Oiling watches 124

Outside drop, close 120

Out of angle 102-120

Overcoil 58,93,94

Pallet and tooth, width of 45

Pallet and wheel action 44

Pallet fork 40,41, 131

Pallet impulse face 40

Pallet letting-off corner 40

Pallet locking face 40

Pallets 14,40, 131

adjusting of 100

circular 42, 46

discharging 40

190

A PRACTICAL COURSE IN HOROLOGY

PAGE

equidistant 41, 46

lift on 105

receiving 40

safety action 45, 110

semitangental 42

Pallet stone 40

Pinion

depthing too deep 62, 64

depthing too shallow 62, 64

minute 28

too large 62, 63

too small 62, 64

Pinions 13,32,62,65,85

Pinning of balance spring

at the collet 143

Pitch

circle 32

circular 33, 37

diameter 32, 35

diametrical 33

Pivot, conical, turning a 77

Pivoting 85-87

Pivot polisher 77, 79

Pivots, making and turn-
ing 74,77

Poise error, analysis of 135

Poise of the balance 134-136

Polishing cone pivot 80

Position adjusting

140-150, 159

PAGE

Practical work of adjust-
ing 159

Principles of gearing 33

Problems in adjusting

escapement 117

isochronism 168

position 161

temperature 170

Ratchet tooth escape

wheel 39

Receiving pallet 40

Regulating and timing 171

Regulator pins, adjust-
ment of 146

Remaining lock 111

Resistance to unlocking

19, 42, 103
Roller, single and double

41,49,50
Roller and fork action_47, 100

Roller axis 82,83

Roller jewel 40,41,49

Roller table 40

Rounding-up tool 66

Safety lock tests 110

Screws, timing 171

INDEX

191

PAGE

Setting jewels (^

Setting staff in cement

chuck 82

Semitangental pallets 42, 110

Shake 105, 120

Short fork 117

Slide 102,113

Slot, fork 40

Springs

balance 52, 88-99

Breguet, fitting 90

flat, fitting 89

Staff, balance 7?>

Staff, in cement chuck 82

Staff making, common er-
rors in 83

Staff, measuring for 75

Staff, turning 76, 83

Staking tool 65

Steel, preparation of 76

Teeth in escape wheel 43

Temperature, adjustment

to 156-158,170

Temperature error _ 52-54, 157

Terminals, theoretical 59

Testing escapement

draw 104

drop 104

PAGE

droplock 101

lift 105

safety locks 110

shake 105

Theoretically correct es-
capement 114

Theoretical terminals 59

Theory of demagnetiza-
tion 137

Third wheel and pinion 16

Three positions, adjust-
ment to 161

Three-to-one roller action 47

Timing and regulating 171

Timing screws 171

Timing washers 171

Toe of tooth 39

Tooth, width of 35-46

Torsion 152, 154

Train 13

dial 15, 28-31

fast 19

main 15, 129

problems 62-67

repairing 65

slow 18

wheel 129

wheel, stretching 6

Tripping error 117-119

192

A PRACTICAL COURSE IN HOROLOGY

^ PAGE

Unlocking and impulse ac-
tions 47

Unlocking resistance, an-
alysis of 42

Use of regulator pins in

adjusting 146

Vibrating a balance spring 89

Vibrations 88

Vibrator 89,90

Washers, timing

171

PAGE

Watch-cleaning machine _ 125

Wheel and pallet action

44,100
Wheel and pinion prob-
lems 62

Wheel

reducing 66

stretching 65

Wheels and pinions 13

Wheel work 13-31

Width of crescent 50

Width of pallet 45

Width of tooth 35,46

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Books

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "A practical course in horology"

See other formats