bever ly je rol d

The French Time Devices Revisited

Much disparity exists among the metronome marks derived from the tempo numbers
for early eighteenth-century French time devices. While some are reasonable, others are
implausibly rapid. A newly discovered source, which offers both Raoul Auger Feuillets
numbers for various forms and a drawing of the pendulum device for which they were
intended, solves the mystery of the conflicting numbers. Because only a clockwork
mechanism can measure fractions of seconds, his numbers had to measure pendulum
lengths (the simpler and most frequent form of measurement). A comparison of his
numbers with those for the same dance forms from the two sources with consistently
extreme tempos indicates an almost exact correlation when all are measured according
to pendulum length, instead of the presumed sixtieths of a second.

For some eighty years, the tempo numbers for French dance music and certain vocal pieces,
derived from time-measuring devices and presented principally in a few French writings
from 1696 to 1762, have been a topic of lively discussion.1 When converted into metronome
marks, many of these numbers for the same form are significantly inconsistent. Although
the very rapid tempos have often been considered valid, the conflict between these and the
other much slower tempos for the same forms has not been explained adequately. Why
are the numbers attributed to Joseph Sauveurs clockwork measurement system (1701) by
Michel LAffilard (1705) and Louis-Lon Pajot, comte dOnzembray (1732) completely
out of range from the one tempo number that Sauveur himself supplied and also from
those of tienne Louli (1696)? Why do Jacques-Alexandre de La Chapelle (1737) and
Henri-Louis Choquel (1762) provide some numbers of very modest speed, but others
that are extraordinarily rapid? Because all of these writers numbers are readily available
in the modern literature (note 1), they will not be repeated again, except when relevant
to material in a recently discovered source that illustrates and describes the pendulum
designed by the Paris dancing master Raoul Auger Feuillet (d.1710). His numbers for
various dance forms provide the most accurate and plausible large body of information
to date about tempo of the period. At this time, two principal forms of measurement
existed: one based on pendulum length in inches (pouces) and the other on sixtieths of
a second (tierces). The latter, however, requires a complex clockwork mechanism. It was
the confusion between these two measurement systems that produced unusually rapid
tempos in two sources. The disparities in the other two sets of numbers can be attributed
to other factors. Throughout this article, the term metronome, identified by an M, refers
only to the modern device, whose mechanism bears no relation to its forerunners.

See, for example, Eugne Borrel, Les indications mtronomiques laisses par les auteurs franais du XVIIIe
sicle, Revue de musicologie 9 (1928), 149-153; Ralph Kirkpatrick, Eighteenth-Century Metronomic Indications,
Papers of the American Musicological Society (1938), 30-50; Hellmuth Christian Wolff, Das Metronom des LouisLon Pajot 1735, in: Nils Schirring, Henrik Glahn, and Carsten E. Hafling (eds), Festskrift Jens Peter Larsen,
Copenhagen: Wilhelm Hansen, 1972, 205-217; Willem Retze Talsma, Wiedergeburt der Klassiker: Anleitung zur
Entmechanisierung der Musik, Innsbruck: Wort und Welt Verlag, 1980; Rebecca Harris-Warrick, Interpreting
Pendulum Markings for French Baroque Dance, Historical Performance 6 (Spring 1993), 9-22; and Klaus
Miehling, Das Tempo in der Musik von Barock und Vorklassik, second edn, Wilhelmshaven: F. Noetzel, 2003.

dutch journal of music theory, volume 15, number 3 (2010)

TvM_15_#3_nov_2010_7.indd 169

169

15-11-2010 12:27:51

the french time devices revisited

Figure 1
Louli, Chronomtre.

Measurement by Pendulum Length

Loulis chronomtre (Figure 1), a simple pendulum, stood
over six feet high. As Louli specifies, the measurement is
according to the pied universel 33.12 cm. with a pouce (royal
French inch) of 27.6 mm. Thus the pendulum length for one
second of time is just slightly over 36 pouces, equivalent to
the English 39.1 inches. The formula for a metronome mark
is number360of pouces . Despite the devices lack of graduated scaling,
three of his numbers for four incipits of pieces from sonatas
by an unknown composer (Example 1) produce plausible
metronome derivations.2 An exception is Example 1b, whose
pendulum length of 8 pouces has vibrations too rapid for the
eye to measure accurately with ease, and may be a misprint.
The shortest length for a piece using Feuillets pendulum, to
be discussed below, is 24 pouces.

After visiting Paris in 1715-1716, the German architect,
librettist, and intellectual Johann Friedrich Armand von
Uffenbach returned to Frankfurt with a Feuillet chronomtre
(Figure 2), which had tempo numbers for seventeen French
dances and Entres (Figure 3) affixed to the bottom of its post.
As the journal of his travel experiences states: Eine Maschine
den Tact in der Musik anzugeben, von der Erfindung des Hr
Feuillets zu Paris.3 In 1728, Uffenbach gave a presentation
about this device (included in his papers) to a learned society
in Frankfurt.4 According to his text, Feuillet invented the
chronomtre at the behest of King Louis XIV because he could
not hear any harmony (Stimmen) among the instruments
in music performances, particularly in operas, and could not
bear disharmony or disorder. Because there was perpetual
strife between the dancers and the opera orchestra concerning
whether a ballet entre or other song was played quickly or
slowly enough, the inventor constructed a small device by
which the beat or tempo could always be the same, and thus
guide both the orchestra and the dancers on stage. It consists
of a 2-inch square post that is 5 feet long and marked with
a scale of unevenly spaced sections (thus an improvement
over Loulis device, which did not use graduated scaling).
When the bob moves in front of the circular mirror on
the post, it casts a shadow that enables the eye to grasp the
beat more precisely. Uffenbachs drawing in Figure 2 shows
front and side views of a simple pendulum with graduated
tienne Louli, lments ou principes de la musique, Paris: Ballard, 1696; facsim. edn, Geneva: Minkoff, 1971,
86ff. The note value placed above the pendulum length in pouces designates the beat unit.
3 Jrgen Kroemer, Le Cronomtre de Monsieur Feuillet: Absolute Tempoangaben eines barocken
Tanzmeisters, sterreichische Musikzeitung 56/7 (2001), 23-28.
4 D-Gs, Cod. Ms. Uffenbach 13/II, 249-254. Figures 2 and 3 from this manuscript are reproduced with the kind
permission of the Niederschsische Staats- und Universittsbibliothek Gttingen. Uffenbachs handwriting
is in old German script, a transcription of which is in the Appendix at the end of this article.
2

170

TvM_15_#3_nov_2010_7.indd 170

15-11-2010 12:27:52

dutch journal of music theory

Example 1
Louli, Sonata incipits.

a.

c.

b.

d.

Incipit
a.
b.
c.
d.

Two beats lents

Four beats lgrs
Trs lents
A final movement

Time signature Beats/ bar

Loulis number

Metronome mark

C-barr
C
3/2
6/4

40
8
30
16

57
127
66
90

2
4
3
2

scaling. Therefore, the tempo numbers cannot be in the tierce (sixtieths of a second) time
measurement proposed today5 because this requires a clockwork mechanism.
Quoting from the French text included with the chronomtre, Uffenbachs commentary
explains the crescents surrounding the number for each dance form in Figure 3. Except for
one omission, the beat unit corresponds to the system described by Michel LAffilard (1705):6

No crescents = one beat/bar

A crescent above = two beats/bar
A crescent on the left = three beats/bar
Crescents above and below = four beats/bar
Crescents on both sides = six beats/bar (in LAffilard only)

Without a clockwork mechanism, Feuillets numbers must be interpreted as pendulum

lengths instead of tierces. Those in Figure 3 produce reasonable metronome derivations
(Table 1). Corresponding almost exactly to Feuillets numbers in Table 1 are the six for
dances in an early eighteenth-century manuscript of choreographies in Feuillet notation,
which likewise utilize crescents to indicate the beat unit (Table 2).7 The numbers appear to
be contemporaneous with the manuscript and may be from the same hand as the dances.
5 Kroemer, Le Cronomtre, 25f. and Miehling, Das Tempo, 59.
6 Michel LAffilard, Principes trs-faciles pour bien apprendre la musique, fifth edn, Paris: Christophe Ballard,
1705; facsim. edn, Geneva: Minkoff, 1971. Directions for interpreting the beat units are on folding plate II
(inserted by p.55). His instructions are also reprinted in Rosamond E. M. Harding, Origins of Musical Time
and Expression, London: Oxford University Press, 1938, plate 10.
7 F-Po ms. 817. See Harris-Warrick, Interpreting pendulum markings, 21f. For Feuillets Sarabande, the
number is uncertain. Of the four possibilities, 38 duplicates that specified in Figure 3 for this dance. This
manuscript is described by Meredith Ellis Little and Carol G. Marsh, La Danse Noble: An Inventory of Dances
and Sources, Williamstown, Mass.: Broude Brothers, 1992, 132f.

171

TvM_15_#3_nov_2010_7.indd 171

15-11-2010 12:27:52

the french time devices revisited

Figure 2
Uffenbachs drawing of Feuillets chronomtre.

Figure 3
Feuillets tempo numbers.

Since the highest number of the chronomtre described by Uffenbach is 60, it cannot be
an exact replica of Feuillets, for his numbers extend to 90. Nevertheless, its form had to
be similar. Uffenbach probably purchased it from the Atelier chez Feuillet, continued by
Jacques Dezais after Feuillets death, which would have found a more ready market for
a device of less imposing dimensions than the one Feuillet needed for his own use with
dancers. Because it is difficult to gauge tempo visually by a rapidly moving pendulum
lacking an audible signal, it was advantageous to have one of sufficient size to measure
a slow compound metre, as in the Chique lente in Table 1. The French text quoted by
Uffenbach advises the user to subdivide the beat when the number extends beyond the
devices range, as with 74 for the Entre lente. While workable for this dance because
it is in duple metre this approach cannot be used with the compound-metre forms.
172

TvM_15_#3_nov_2010_7.indd 172

15-11-2010 12:27:53

dutch journal of music theory

Table 1
Metronome marks from Feuillets pendulum.
Dance

Time
signature

Beats/bar

Feuillets
number

Metronome
mark

Menuet
Passepied
Gaillarde
Gavotte
Entre vite
Entre lente
Entre lente
Bourre
Rigaudon
Sarabande
Passacaille
Courante
Chaconne
Chique lente
Loure
Gigue vite
Canary

3
3/8
C-barr
C-barr
C-barr
2
C
2
2
3
3
3/2
3
6/4
6/4
6/4
6/8

1
1
2
2
2
2
4
2
2
3
3
3
3
2
2
2
2

48
40
40
37
37
74
37
30
27
38
36
36
24
90
78
30
26

52
57
57
59
59
42
59
66
69
58
60
60
73
38
41
66
71

Table 2
Metronome marks from Feuillets numbers in scores.

Dance
Entre de paysant
Gigue de Mr Feuillet
(gigue de thetis et pellee)
Gigue de Mr Feuillet
(gigue de polixenne)
Entre de Mr Feuillet
Chaconne de Mr Feuillet
Sarabande de Mr Feuillet

Time
signature

Beats/bar

Feuillets
number

Metronome
mark

2
6/8

2
2

30
30

66
66

6/4

30

66

C-barr
3
3

2
3
3

30
24
38

66
73
58

Uffenbach obtained his chronomtre some five years after Feuillets death, so the French
writer probably overlooked the difference between duple and compound metres.
In closing his presentation, Uffenbach observes that this machine not only enables
conformity between dancers and musicians, but also lessens the arguments about correct
tempo. Moreover, it helps those who are not yet strong in keeping a steady beat, thereby
relieving the (loudly audible) time beating (Geklopfe) during the music. The form of
this time beating is clarified by a footnote in an anonymous English translation (1709)
of Franois Raguenets comparison of French and Italian music (1702). In response to
Raguenets remarks about assembling the various elements at the Paris Opra:
How many times must we practice an opera before its fit to be performed; this man begins
too soon, that too slow; one sings out of tune, another out of time; in the meanwhile the
composer labors with hand and voice and screws his body into a thousand contortions and
finds all little enough to his purpose.
173

TvM_15_#3_nov_2010_7.indd 173

15-11-2010 12:27:53

the french time devices revisited

the translator observes:

Some years since, the master of the music in the opera at Paris had an elbow chair and desk
placed on the stage, where, with the score in one hand and a stick in the other, he beat time
on a table put there for that purpose so loud that he made a greater noise than the whole
band, on purpose to be heard by the performer. By degrees they removed this abuse from
the stage to the music room [probably the orchestra pit], where the composer beats the time
in the same manner and as loud as ever.8

An accident while beating time with a rod led to Jean-Baptiste Lullys premature demise
in 1687 when a blow to his toe became infected. Nevertheless, to the chagrin of critics,
distracting conducting continued at the Paris Opra for much of the eighteenth century.
According to Jean-Jacques Rousseau (1768), the French did not use a roll of paper for
beating time, as commonly done elsewhere, but a large baton of hard wood, which was
struck forcefully to be heard from afar.9
The musicien inconnu La Chapelle, too, used pendulum measurement for many
incipits of unknown pieces in his primer, but the metronome marks derived from his
numbers are widely disparate.10 While some are plausible, others are so extremely fast
as to have no relation to the others. La Chapelle provides no beat unit for any of his
numbers, and it is likely that the extreme tempos should have a smaller beat unit than
has been calculated. Because he applies the time signature 2 indiscriminately for all
forms of duple movement (even the allemande, to which early sources nearly always
assign four slow beats and a signature of C), the beat unit is uncertain. According
to writers such as Jacques Hotteterre (1719), the C-barr signature, for example,
can have either two slow or four faster beats (depending on the pieces texture and
predominating note values).11 In 1767, the critic Pascal Boyer observed that time
signatures were never intended to tell the musician what to do with his body: When
beating the measure of two beats, several music masters make four hand movements,
while others make eight motions for the measure of four beats, etc., without anyone
ever accusing them of not knowing how to beat time.12 A further complicating factor
is that some composers (such as Jean-Philippe Rameau) did not apply the signatures
in the conventional manner.
Using an incorrect beat unit with La Chapelles numbers, mainly those in duple metre,
is what has produced untoward tempos. On the other hand, a crotchet beat unit is often
satisfactory when the signature is 3. And for the signature of 3/2, La Chapelle includes
an incipit of two voices comprising crotchets and minims, which is assigned a moderate
tempo of minim = M 54. A Rondeau in compound-metre 6/8, composed of crotchets and
quavers, is marked as dotted crotchet = M 66.13 Thus the extreme tempos occur principally
8

9
10
11
12
13

Franois Raguenet, Parallle des Italiens et des Franais en ce que regarde la musique et les opras, Paris: Jean
Moreau, 1702; facsim. edn, Geneva: Minkoff, 1976, 96f. English translation in A Comparison between the
French and Italian Musick and Operas, London: W. Lewis, 1709, 42f. Reprinted in The Musical Quarterly 32/3
(1946), 428f.
Jean-Jacques Rousseau, Dictionnaire de musique, Paris: Vve. Duchesne, 1768, Baton de mesure.
Jacques-Alexandre de La Chapelle, Les vrais principes de la musique, Paris: lauteur, la veuve Boivin, 1736-1752,
vol. 2, 41-56. His examples are supplied in Miehling, Das Tempo, 85-91.
Jacques Hotteterre, LArt de prluder, Paris: lauteur, Boivin, 1719; facsim. edn, Geneva: Minkoff, 1978, 57.
Pascal Boyer, Lettre Monsieur Diderot sur le projet de lunit de clef dans la musique. Et la rforme des mesures,
Amsterdam; Paris: Vente, 1767, 52-54, note.
La Chapelle, Les vrais principes, Leons deux parties, voix egalles, vol. 3, 1-3. For examples, see Miehling,
Das Tempo, 90, nos. 43, 45.

174

TvM_15_#3_nov_2010_7.indd 174

15-11-2010 12:27:53

dutch journal of music theory

with duple metre, indicating that the probable beat unit for most of these pieces should
be smaller than assumed today.
Another writer using pendulum-length measurement was the attorney Choquel,
whose book includes numbers for five dance forms and eleven pieces from sacred and
secular vocal works.14 While the dances have extreme tempos, most of the vocal pieces
are moderate. For example, Si des Galants de la ville (signature of 2) from Jean-Jacques
Rousseaus Le Devin du village is assigned a pendulum length of 24 pouces, or minim = M
73. The vocal line moves in crotchets, accompanied by quavers in the upper strings, and
the pieces marking of Gai is the fastest one in Choquels examples.15
One of Choquels vocal pieces with a questionable tempo an excerpt in duple metre
from an unnamed motet by Michel-Richard de Lalande lacks a beat-unit indication.16
Two other vocal pieces with unusually rapid tempos are based on dance forms: an Air
en Rondeau from Jean-Baptiste Lullys opera Thse, specified to be a gigue; and a duet
having a Mouvement du Menuet.17 In sum, Choquels numbers are reasonable for eight
vocal pieces, questionable for three vocal pieces, and extreme for five dance forms. We
may find an explanation below.

Measurement by Time
The other writers offering many tempo numbers are the court singer LAffilard and
the scientist Pajot. Unlike those of La Chapelle and Choquel, their numbers seem fairly
consistent within each set of pieces, but are much more rapid than contemporary verbal
descriptions imply. They purport to follow a scaling based on sixtieths of a second (or
tierces), as presented by the mathematician Joseph Sauveur (1701) for his chomtre.
Sauveur furnished no diagram of his device, but it had to have included a clockwork
mechanism to measure fractions of seconds. Sauveurs contemporary Chapotot, a Paris
instrument maker, built chomtres, and one survives in the collection of the Paris
Conservatoire des Arts et des Mtiers. Since Sauveurs pendulum cord was environ de
8 pieds (106 English inches) in length, the massive device could not have been widely
used. He provides a tempo number for just one piece Allons, allons, accourez tous from
Lullys Atys (Example 2).18 With a conversion formula of M = number360oof tierces , his number of
70 translates to a plausible minim = M 51.To achieve this tempo with Loulis chronomtre,
he specifies a pendulum length of 42 pouces, which produces M = 55.5.19 The absence of
a graduated scale in Loulis pendulum accounts for some discrepancy in metronome
derivations. Sauveurs device, too, might not have been quite accurate, or he may have
used one of the differing measurements for the pied.

14 Henri-Louis Choquel, La musique rendue sensible par la mchanique, second edn, Paris: Christophe Ballard,
1762; facsim. edn, Geneva: Minkoff, 1972, 115-213.
15 Choquel, La musique, 180ff.
16 Choquel, La musique, 201f.
17 Choquel, La musique, 186ff., 207ff.
18 From Jean-Baptiste Lully, The tragdies lyriques in facsimile, New York: Broude International, 1998-2007.
Reproduced with kind permission.
19 Joseph Sauveur, Principes dacoustique et de musique:ou Systme gnral des intervalles des sons, [Paris: s.n.,
1701]; facsim. edn, Geneva: Minkoff, 1973, 49f.; also in Joseph Sauveur, Collected Writings on Musical Acoustics
(Paris 1700-1713), ed. Rudolf Rasch, Utrecht: The Diapason Press, 1984, 147f. The latter (p. 40) includes a
photograph of the Chapotot chomtre at the Paris Conservatoire des Arts et des Mtiers. Sauveur measures
Lullys piece also in twelfths of a second (14); the conversion formula is M = 720/n.

175

TvM_15_#3_nov_2010_7.indd 175

15-11-2010 12:27:54

the french time devices revisited

Example 2
Lully, Atys, Allons, allons, accourez tous, Act 1, Scene 2.

Four years later, LAffilard attributed tempo numbers for various pieces in his Principes
trs-faciles pour bien apprendre la musique to Sauveurs system.20 These astonishingly rapid
tempos, which differ greatly from Sauveurs own tempo number, appear in a primer for
beginning vocal pupils. Since vocal agility takes many years to develop and never attains the
speed of which instruments are capable, this requires further investigation; for example:
The text of a Gigue in 3/8 (Example 3a), whose tempo number of 31 per bar is
translated as M 116, cannot be enunciated at this tempo.
For slow forms such as sarabande and courante, LAffilards numbers do not permit
an expressive performance. A tempo of crotchet = M 106 is assigned to his Passacaille
(Example 3b), but it contains successive semiquavers with separate syllables; his
previous edition marks it as Fort gravement. The text is a lament of spurned love: How
many tears have I shed without moving you?
For the four pieces that LAffilard identifies as la mesure six tems graves, the
metronome marks derived range from 120 to 150 per crotchet, and do not qualify
as very slow. When each crotchet = M 150, the correct beat unit has to be two beats
of compound metre. Yet he specified six very slow beats per bar, as spelled out by his
system of enclosing the tempo number with a crescent on both sides.21
LAffilard called his pieces appropriate for (social) dancing, which implies moderate
tempos. The abundant ornamentation, too, requires adequate time for its execution.
Example 3a
LAffilard, Gigue.

Example 3b
LAffilard, Passacaille.

20 LAffilard, Principes, 52-151.

21 LAffilard, Principes, 105, 89, 125-138. Talsma, Wiedergeburt, 154-169 and Miehling, Das Tempo, Anhang 2,
present LAffilards pieces in modern notation.

176

TvM_15_#3_nov_2010_7.indd 176

15-11-2010 12:27:54

dutch journal of music theory

In 1974 Erich Schwandt proposed that the scaling of LAffilards pendulum differed
from Sauveurs, thus making modern translations of LAffilards numbers twice too
fast.22 With some exceptions, Schwandts corrected numbers correspond more closely to
contemporary descriptions of the dance forms.23 Yet there may be a way to bring nearly
all of LAffilards numbers within a plausible range. While he believed that he was using
Sauveurs system, he was not a mathematician. The numbers supplied are more consistent
with Loulis scaling for pendulum lengths in pouces. Table 3 provides metronome marks
for LAffilards pieces as derived from measurement in both tierces and pouces.
Table 3
LAffilards numbers measured in Tierces and Pouces.

Beats/bar

Metronome
mark from
tierces

Metronome
mark from
pouces

C
2
2
2
2
2

4
2
2
2
2
2

120
120
120
120
90
106

66
66
66
66
57
62

50
45
74 ?
40

3/2
3/2
3/2
3/2

3
3
3
3

72
80
49
90

51
54
42
57

PAR LE TRIPLE SIMPLE

Sarabande en Rondeau
Passacaille
Chaconne
Menuet

42
34
23
51

3
3
3
3

3
3
3
1

86
106
157
71

56
62
75
50

PAR LE TRIPLE MINEUR

Passepied
Gigue
Air fort leger

42
31
31

3/8
3/8
3/8

1
1
1

86
116
116

56
65
65

A SIX TEMS GRAVES

Leon
Sarabande
Marche en Rondeau
Air grave en Rondeau

24
27
24
30

6/4
6/4
6/4
6/4

6
6
6
6

150
133
150
120

73
69
73
66

A SIX TEMS LEGERS

Canaries en Rondeau
Menuet
Gigue

34
48
36

6/8
6/8
6/8

2
2
2

106
75
100

62
52
60

LAffilards
number

Time
signature

A DEUX TEMS
Marche
Gavotte
Rigaudon
Bourre
Pavane
Branle en Rondeau

30
30
30
30
40
34

PAR LE TRIPLE DOUBLE

Sarabande tendre
Air tendre
Air, fort grave
Courante

22 Erich Schwandt, LAffilard on the French Court Dances, The Musical Quarterly 63 (1974), 395.
23 Erich Schwandt, LAffilard, in: Stanley Sadie and John Tyrrell (eds), The New Grove Dictionary of Music and
Musicians, second edition, London: Macmillan, 2001, vol. 14, 109.

177

TvM_15_#3_nov_2010_7.indd 177

15-11-2010 12:27:54

the french time devices revisited

With one possible exception, none of the tempos derived from pendulum lengths is
unusual. They are, in fact, quite similar to Feuillets. One of LAffilards numbers is out
of range from the rest: the 74 for an Air, fort grave (Example 4), which is a reasonable
tierce number for this piece.24 Perhaps the tempo measurement was first undertaken with
Sauveurs system, and then converted to pendulum-length measurement, for Sauveurs
device must have been too large and expensive to find a market. In the changeover, the
number 74 was overlooked. Because practicing musicians rarely had access to more
than the most rudimentary general education, it is unlikely that LAffilard prepared the
purported tierce numbers himself. More probably, he enlisted the aid of a mathematician,
who then failed to communicate the change to him. Louli, who may have been the only
musician capable of catching the error, had died three years earlier.
Example 4
LAffilard, Air, fort grave.

LAffilards misattribution of his numbers to Sauveurs tierce measurement might explain

why most of Choquels numbers for vocal pieces are reasonable, while those for dance
forms (which include two other vocal pieces) are excessively fast. For the dance forms
(Gavotte, Rigaudon, Menuet, Passepied, and Gigue), Choquel simply converted LAffilards
numbers from the assumed tierces into pendulum pouces, making slight adjustments.
The last set of numbers is found in Pajots Description et usage dun mtromtre, where he
calls his machine an improvement of Loulis chronomtre because it is measured in parts
of a second instead of pendulum pouces, uses an aural signal to identify the beginning and
last part of each pendulum swing, and has a graduated scale.25 Pajots Table of pendulum
lengths (partially supplied in Figure 4) comprises those for the different durations of
vibrations from demi-tierce to demi-tierce up to 180 demi-tierces, or a second and a half ,26
using these values:
Pied [foot 331 mm.].
Pouce [royal French inch], the twelfth part of a pied.
Ligne, the twelfth part of a pouce.
Point, presumably the twelfth part of a ligne.
The fundamental measurements are as follows:

24 LAffilard, Principes, 77ff.

25 Louis-Lon Pajot, comte dOnzembray, Description et usage dun mtromtre, ou machine pour battre les
mesures & les temps de toutes sortes dairs, in: Histoire de lAcadmie Royale des Sciences, 1732, Paris, 1735,
Mmoires, 182-196.
26 Pajot, Description, 183: & nous y joindrons une Table de toutes les longueurs du Pendule, en pieds, pouces,
lignes & points, pour les diffrents dures des vibrations de demi-tierce en demi-tierce jusqu 180 demitierces, ou une seconde & demie.

178

TvM_15_#3_nov_2010_7.indd 178

15-11-2010 12:27:55

dutch journal of music theory

Everyone knows that an hour is divided into 60 minutes ['], 1 minute into 60 seconds [''],
and 1 second into 60 tierces ['''] or 120 half-tierces; this will give us a sufficiently small division for what we propose. It is also known that a pendulum must have a length of 3 pieds
and 8 lignes, for each vibration to last a second or 60 tierces.27

His full chart of pendulum lengths runs from to 90 tierces, and its unprecedented
mathematical exactitude is the most probable reason that his work was accepted by the
Acadmie Royale des Sciences. The column headed Nombre des demi-tierces contains
tierces, with the demi-tierces inserted between each tierce. Thus the number 60 in this
column requires a pendulum length of 3 pieds and 8 lignes, the correct length for a
second.
Figure 4
Pajot, Table for pendulum lengths (fragment).

27 Pajot, Description, 187f.: Tout le monde sait quune heure se divise en 60 minutes, 1 minute en 60
secondes, et 1 seconde en 60 tierces ou 120 demi-tierces; cela nous donnera une division suffisamment
petite pour ce que nous proposons. On sait aussi que la longueur que doit avoir un Pendule, pour que
chaque vibration soit dune seconde ou de 60 tierces, doit tre de 3 pieds 8 lignes et demi.

179

TvM_15_#3_nov_2010_7.indd 179

15-11-2010 12:27:55

the french time devices revisited

Figure 5
Pajot, Mtromtre.

180

TvM_15_#3_nov_2010_7.indd 180

15-11-2010 12:27:56

dutch journal of music theory

Pajot describes his machine (Figure 5, which includes a simple pendulum in between
front and side views of his own device) as follows:
The two vertical pieces A, B, and C, D are each about five feet in length . On top of these
two pieces is a pendulum E, whose beats of the bob are heard distinctly; thus one hears the
beginning and end [part] of each vibration. There are holes to mark 76 demi-tierces; in
other words, from 30 to 68 tierces.28

In his chart of tempo numbers for pieces from Lully, Pascal Collasse, Andr Campra, AndrCardinal Destouches, and Jean-Baptiste Matho (Figure 6), the third column supplies the
time signature; the fourth, the number of beats per bar; the fifth, the number of tierces
per bar; and the sixth, the number of tierces per beat. As with the tierce interpretation of
LAffilards numbers, Pajots numbers are amazingly rapid.
Figure 6
Pajot, Chart of tempos.

28 Pajot, Description, 184ff.: Les deux montants verticaux A,B, & C, D, ont chacun environ 5 pieds de hauteur
. Sur ces deux montant est une Pendule E, dont les battements du rocher se sont entendre distinctement,
ainsi on connoit par loreille le commencement & la fin de chaque vibration. lon a fait des trous pour
marquer 76 demi-tierces, savoir depuis 30 jusqu 68 tierces.

181

TvM_15_#3_nov_2010_7.indd 181

15-11-2010 12:27:57

the french time devices revisited

According to Pajots text, his machine has an aural signal to mark both the beginning of
each pendulum swing and its return (a period). A period lasting one second (60''') would
therefore have audible signals spaced a half second apart (or M 120). For the fastest tempo
on his machine (30'''), these signals would be at quarter-second intervals (or M 240).
But it is doubtful that technology existed for attaining an audible signal at such speed.
Moreover, the ear cannot distinguish individual components moving so rapidly, making
the machine useless for determining tempo. Thus Pajots tierce numbers for pieces in
Figure 6 do not appear to correlate with his machines description.
After Loulis death in 1702, Pajot acquired his chronomtre. In 1696, Louli noted
that he had consulted with musicians who had performed under Lully, after which he
calculated tempo numbers for various pieces.29 These numbers may have been inserted
into Loulis personal copies of scores in his extensive library, which was apparently
dispersed after his death, or they may have existed in a master list. No trace of them has
come to light. When obtaining Loulis chronomtre, the collector Pajot may also have
acquired some of his library or a list of his tempo numbers. All of the pieces for which
Pajot provided tierce numbers in Figure 6 were composed during Loulis lifetime. As has
been proposed, these numbers may have derived from Loulis missing ones.30
Just as LAffilard was not a mathematician, Pajot had no music credentials, as can be
verified by certain items in his chart. For instance, the second Air des songes funestes
from Lullys Atys (Act 3, Scene 4) has a time signature of 3/2.31 Yet Pajot divides the bar
into two parts (thus 6/4) instead of three.
Even though Pajots chart specifies that Les Dmons (actually Feste Infernale; Act
4, Scene 3) from Lullys Alceste has 4 temps, he divides the C signature into two parts,
instead of four. Therefore, he did not himself provide the four-beat description. This
signature conveyed four beats, normally slow unless indicated otherwise. The designation
4 temps likely derives from a notation in a list that Louli compiled, for it would be
unnecessary in the edition itself. Since the other pieces in this scene have different time
signatures, it served to identify the one intended.
An incipit for the Loure from Collasses Thetis & Pele in Pajots chart is included
in Hotteterres description (1719) of the 6/4 signature. Calling its tempo grave, he
recommends four unequal beats (two minim/crotchet units).32 Since Pajot implausibly
assigns the Loure the same tempo as the rapid Gigue, the tempo number itself is probably
incorrect. Further errors or questionable aspects of Pajots table include:
A Gigue from Lullys Amadis is misattributed to Collasse.
The Menuet from Campras lEurope galante has an incorrect time signature of 2.
Lullys Ftes de lamour et de Bacchus has no Chaconne des Arlequins. Its purported
number 68 for a full bar measured in tierces would produce a tempo almost twice as
fast as Feuillets chaconne.
Although Pajot lists a Divinits de la terre from Lullys Perse, none exists in this
opera. Scholars have inferred that it must be the Entre de divinitez infernales, but
29 Louli, lments, 88.
30 See Patricia M. Ranum, Mr de Lully en trio: Etienne Louli, the Foucaults, and the Transcription of the
Works of Jean-Baptiste Lully (1673-1702), in: Jrome de La Gorce and Herbert Schneider (eds), Jean-Baptiste
Lully: Actes du colloque = Kongressbericht: Saint-Germain-en-Laye, Heidelberg 1987, Laaber: Laaber-Verlag,
1990, 314.
31 For this piece, Wolff, Das Metronom, 216, and Miehling, Das Tempo, 80, select the preceding chorus, also
in 3/2.
32 Hotteterre, LArt de prluder, 59. Until corrected in Miehlings second edition of Das Tempo (81), writers
have cited a different piece from this opera, which, however, is not a Loure, but carries the expression mark
Lour.

182

TvM_15_#3_nov_2010_7.indd 182

15-11-2010 12:27:57

dutch journal of music theory

this is speculative. Perhaps Pajot listed the wrong piece or opera.

Multiple possibilities exist for Les Dmons from Lullys Psych: the Prlude in Act 4,
Scene 1, where the demons enter and begin to terrify Psych; the next piece (Scene 2)
with the three Furies and Psych; the Air des Dmons that follows; and the Prlude to
Act 4, Scene 3, which involves the three Furies, two Nymphes of Acheron, and Psych
(writers today have chosen the latter).
For the first Air des Songes funestes from Lullys Atys, different possibilities have been
presented today.33
The Courante nearly always had a time signature of 3/2, so the beat unit of Mathos
unidentified Courante is probably a minim.
These discrepancies indicate that the chart was not completely Pajots own work. It is more
likely that he compiled it from Loulis numbers in a list incorporating abbreviations and
notations. This list may have comprised nothing more than a title for each piece and its
pendulum length. Using this thesis, the last column in Pajots chart (Figure 6) contains
Loulis numbers. When this column is blank, Loulis number includes an entire bar in
triple metre and is found in the preceding column. The one exception Le Printemps
de Phaton may have an incorrect time signature (several possibilities fit this title), for
duple metre could be halved to obtain a number for the last column. Pajot then misread
Loulis numbers as tierces, instead of pendulum pouces. He calculated the number of
beats in each bar and the resulting number of tierces. But in some instances he may have
misinterpreted the beat unit. Like us, he sometimes had to guess which piece Louli
meant. Moreover, handwriting can easily be misread. Table 4 provides Pajots original
number for a beat (or bar when indicated), and the metronome marks derived from both
tierce and pouce measurement.
Pajots chart appears to have been prepared independently of his own machine, which,
if its description is accurate, would have produced audible signals too rapid to be useful
in most cases. While he clearly intended to achieve tierce measurement, his machine
may actually have been based on pendulum length. He presents himself as building on
Loulis work, and the highest number on his machine is nearly the same as on Loulis
chronomtre.
In contrast to the questionable identity of some free forms in Pajots chart, that of the
dance forms is more certain. When the numbers from LAffilard, Pajot, and Feuillet are all
interpreted as pendulum lengths, as Feuillets must be, the metronome derivations for each
dance form are remarkably similar (Table 5). Besides providing reasonable tempos, pouce
measurement removes the disparity found among some of the dances when measured in
tierces. For example, the pace of LAffilards Sarabande in 6/4 measured in pouces is not
greatly faster than the other Sarabandes; with tierce measurement; on the other hand, the
metronome marks are 72, 86, and 133. While early sources define the Chaconne as just
somewhat faster than the Passacaille, tierce measurement produces M 157 for the former
and 106 for the latter.
None of the numbers in Table 5 should be regarded as a fixed tempo, but as an
approximation to be adjusted up or down according to the pieces texture. Some dances
existed in multiple forms: for example, Jean-Jacques Rousseau describes the Gavotte
as ordinarily graceful, often gai; also sometimes tender and slow.34 Choquel makes an
interesting point when observing that it would be better to write the Menuet in 6/4
instead of 3, because the Pas de Menuet comprises two bars of 3, each of which has one
step. Thus the Matres Danser beat the Menuet in two one beat for each bar of 3, which

33 See Miehling, Das Tempo, 79; and Wolff, Das Metronom, 216.
34 J.-J. Rousseau, Dictionnaire, Gavotte.

183

TvM_15_#3_nov_2010_7.indd 183

15-11-2010 12:27:57

the french time devices revisited

Table 4
Pajots numbers measured in Tierces and Pouces.

Beats/bar*

Metronome
mark from
tierces

Metronome
mark from
pouces

32
34

2
?

112
106

64
62

68 (full bar)
37
45
63
32

??
2
??
??
3

106
98
80
58
112

44
59
54
45
64

30
38
48
361/2
68 (full bar)

2
3
4
??
??

120
95
76
100
53

66
58
52
60
44

32
32?

2
?

112
112

64
64

56
45

2
2

64
80

48
54

Campra
Passepied de lEurope galante
Rigaudon de lEurope galante
Menuet de lEurope galante

36 (full bar)
31
51 (full bar)

1
2
1

100
116
71

60
65
50

Destouches
Sarabande dIss
Boure dOmphale
Menuet de Marthsie

49
30
51 (full bar)

3
2
1

72
120
71

51
66
50

Matho
Courante

44

79

54

Lully
Boure de Phaton
La Marie des Ftes de Bacchus &
de lAmour
Le Printemps de Phaton
Gavotte de Roland
Les Dmons de Psich
1.er Air des Songes funestes dAtis
2.d Air des Songes funestes dAtis
Les Dmons du 4.me acte de Proserpine
Passacaille de Perse
Les Dmons dAlceste 4 temps
Les Divinits de la terre de Perse
La Chaconne des Arlequins des Ftes de
Bacchus & de lAmour
Collasse
Gigue dAmadis [actually Lully]
Loure de Thtis & Pele
LOuverture de Thtis & Pele,
Le Commencement
Et la Reprise

Pajots
number

* ? indicates a questionable beat unit, and ?? an uncertain piece.

184

TvM_15_#3_nov_2010_7.indd 184

15-11-2010 12:27:57

dutch journal of music theory

Table 5
A comparison of numbers for dance forms.
Metronome marks derived from numbers interpreted as pouces instead of tierces.
Time signature

Beats/bar

LAffilard

Pajot

Feuillet

Bourre

66

Phaton, 64
Omphale, 66

66

Gavotte

2
C-barr

2
2

66

Roland, 59

Rigaudon

66

LEurope, 65

69

Sarabande

3/2
3
6/4

3
3
2

51
56
69

Iss, 51

58

Passacaille

62

Perse, 58

60

Chaconne

75

73

Menuet

50

LEurope, 50
Marthsie, 50

52

6/8

52

3/8
6/8
6/4

1
2
2

65
60
Amadis, 64

66

Passepied

3/8

56

LEurope, 60

57

Courante

3/2

57

54

60

Canaries

6/8

62

Loure

6/4

Gigue

59

71
?

41

moves too quickly for the hand to beat it comfortably in three.35 His remarks fit with
Table 5s Menuet metronome mark of 50 or 52 for one bar of 3; if the hand had to make
three motions per bar at this speed, it would shortly become fatiguing.
From the similar tempos for each dance form in Table 5, it can be seen that LAffilards
and Pajots numbers were based not on Sauveurs system of tierce measurement, but on
the same pendulum-length measurement that was required for Feuillets device. The
35 Choquel, La Musique, 127: Je crois quil vaut mieux appliquer cette mesure 6 & 4 au Menuet que celle du
triple simple; car le Pas de Menuet absorbant deux mesures trois temps simples, puisque les Matres
Danser font battre le Menuet deux temps dont chacun emporte une mesure triple simple par chaque Pas,
il seroit beaucoup mieux de se runir sur ce point avec eux. La mesure trois temps simples est dailleurs
si presse pour le vrai mouvement du Menuet que la main na pas tout le temps ncessaire pour marquer
chaque temps suivant le triangle que forme cette sorte de mesure.

185

TvM_15_#3_nov_2010_7.indd 185

15-11-2010 12:27:57

the french time devices revisited

many discrepancies in Pajots chart indicate that he constructed it from Loulis missing
pendulum numbers.

Views from Contemporaries

According to Rousseau, Pajots machine succeeded in neither one tempo, nor another.36
Nicolas Framerys comment on Rousseaus article reveals that none of these timemeasuring devices made an impact:
Several have built and proposed different machines, which were aimed at marking and,
in particular, conserving the true tempo of each piece as conceived by the composer; but,
too complicated in their means and too limited [for achieving] their object, none has been
adopted.37

According to Jean-Philippe Rameau, Loulis chronomtre was neglected because of its

difficulty, although it was in other respects an ingenious invention.38 Writing from
the Berlin court in 1752, the flautist Johann Joachim Quantz had never known anyone
who used it.39 The one device that seems to have had practical application (for use with
dancing) was Feuillets. Perhaps more scores with tempo numbers for the dances await
discovery.
Instead of setting tempo with a chronomtre device, the encyclopedist Denis Diderot
suggested in 1748 that composers indicate the amount of time needed to play their piece
in its entirety.40 This method was employed in an autograph manuscript of Lalandes Te
Deum (between 1715 and 1726). At the end of most versets is an annotation with the
performance length, which totals 29 minutes or une bonne demi-heure, written
on the last page. The Te Deum had to fit within the time frame specified by the king.
While the tempo for some movements cannot be established exactly because of different
versions, cuts, optional repeats, or internal metre changes, that for eight movements with
a single time signature and no complicating factors is obtainable.41 All are moderate, and
in keeping with the tempos above from Louli, Sauveur, Feuillet, most of Choquels vocal
pieces, and LAffilards and Pajots numbers when interpreted as pendulum pouces instead
of tierces. Choquels few extreme numbers for dance forms appear to derive from assuming
that LAffilards numbers were tierces. For lack of a beat unit, La Chapelles numbers are
unreliable for scientific inquiry.
Because their standards were not our standards, and their equipment not ours, all of
their numbers must be construed as approximations with a greater or lesser degree of
inaccuracy. They also are subject to the same errors of misprints, mechanical malfunction,
36 J.-J. Rousseau, Dictionnaire, Chronomtre, 99: Il y a une trentaine dannes quon vit parotre le projet dun
Instrument semblable, sous le nom de Mtromtre, qui battoit la Mesure tout seul; mais il na russi ni dans
un tems, ni dans lautre.
37 Nicolas Framery, Chronomtre, in: Dictionnaire mthodique. Musique, Nicolas Framery and Pierre Ginguen
(eds), Paris: Panckoucke, 1791, vol. 1, 280: Plusieurs mchaniciens ont excut & propos diffrentes
machines, qui avoient pour but de marquer & surtout de conserver le vritable mouvement de chaque
morceau, tel quil a t conu par lauteur; mais trop compliques dans leurs moyens, & trop bornes dans
leur objet, aucune na t adopte.
38 Jean-Philippe Rameau, Trait de lharmonie, Paris: Jean-Baptiste-Christophe Ballard, 1722, 158.
39 Johann Joachim Quantz, Versuch einer Anweisung die Flte traversire zu spielen, Berlin: J. F. Voss, 1752, XVII/
vii/46, 261.
40 Denis Diderot, Mmoires sur diffrens sujets de mathmatique, Paris: Durant et Pissot, 1748, 195f.
41 See Lionel Sawkins, Doucement and lgrement: Tempo in French Baroque Music, Early Music 21 (1993),
365-374. The manuscript (F-Pn H400D) is described by Genevive Thibault, Le Te Deum de Lalande:
Minutage de lpoque, Fontes artis musicae 12 (1965), 162-165.

186

TvM_15_#3_nov_2010_7.indd 186

15-11-2010 12:27:57

dutch journal of music theory

and human judgement we see today. Moreover, their lack of metronome training for
musicians led to what we would term rhythmic inaccuracy, which was not entirely
undesirable. As Diderot comments:
Connoisseurs will object to the chronomtre because there are perhaps not four bars in an
air that have the same duration . A musician who knows his art sings or plays more
slowly or less slowly from one bar to another, and even from one beat or quarter-beat to
the following.42

Rhythmic freedom was acceptable for soloists, but created havoc in ensembles. This
explains why leaders had to beat time audibly and why tempos therefore had to be very
moderate in comparison to ours.43 If we had never undergone metronome training from
childhood, we, too, would perform as erratically as Diderot describes. As for the numbers
themselves, it is impossible to obtain an accurate tempo measurement without first
acquiring the ability to maintain a perfectly steady tempo. The dancing master Feuillet
probably had as sound a rhythmic sense as anyone of the period a further reason for
the importance of his numbers. Together with the visual evidence of the pendulum for
which they were intended, these numbers provide the key to interpreting the questionable
or ambiguous numbers of others. With few exceptions, the various sources now present
greater uniformity and plausibility of tempo.

42 Diderot, Mmoires, 193f.: Ils objecteront contre tout Chronomtre en gnral, quil ny a peut-tre pas dans
un air quatre mesures qui soient exactement de la mme dure . Un Musicien qui sait son art chante ou
jou plus ou moins lentement dune mesure un autre & mme dun tems & dun quart de tems celui qui
le suit. Le seul bon Chronomtre que lon puisse avoir, cest un habile Musicien qui ait du got, qui ait bien
l la Musique quil doit faire excuter, & qui sache en battre la mesure.
43 See, for example, J.-J. Rousseau, Dictionnaire, Battre la mesure, 51.

187

TvM_15_#3_nov_2010_7.indd 187

15-11-2010 12:27:57

the french time devices revisited

Appendix
The transcription of Uffenbachs handwritten text in old German script.1*
[249] Nach dieem hergeleenen und an dem Weltmodell erluterten beyden Aufstzen,
zeigte ich der Gesellschafft eine gewie zu der Music dienliche Machine, so ehedeen auf
Befehl Knig Ludwig des XIV. von einem Mitglied der Kniglichen Academie, Feuillet,
erfunden worden, inmaen er sowohl in der Music als insonderheit denen Opern
keine Stimmen der Instrumenten hren, noch einige Ungleichheit oder Unordnung
vertragen konte, weilen es nun unter denen Tntzern und dem Orchester deren Opern
einen immerwhrender Streit gesezet, ob man nehmlich ein Ballet entre oder andern
Gesang nicht geschwind oder langsam genug vorgespiehlet, so hat der Erfinder ein Mittel
ausgesonnen, vermge eines kleinen Instrumentes den Tact Mensur oder Tempo allemahl
einerley zu haben und sich so wohl in dem Orchestere als inter denen Scenen vor die
Tnzer darnach zu richten. Es bestehet aber solches [250] in einem 5 Schu langen und 2
Zoll im Quadrat dicken holzernen viereckenden Stabe, welcher nach Magebung des hin
und her schwenckens eines Senckels, der fornen an einer seidenen dnnen Schnur vibrirt,
viele ungleiche Abtheilungen auf einem aufgeklebeten langen Papier hat, worber ein
meinger viereckender Ring mit einem kleinen Arm, dadurch die Schnur gezogen, hoch
oder niederig gerutschet und gestellet werden kan, da das centrum oscillationis oder die
Lnge des Fadens an dem Perpendicul verndert werden knne, stellet man nun daelbe
hoch oben hin und lt den Faden lang, so giebt es langsame Vibrationes die mehr Zeit
wegnehmen, als wenn der Faden kurtz gelaen wird, durch dieen Unterschied hat der
Erfinder einen Masstab formiren knnen, welcher die accurate Zeit eines Tacts, er seye
lang oder kurtz, bestimmen kan. Die uere Gestalt von dem ganzen Werck kan man
aus beygesezter Zeichnung abnehmen, wo a, b der lange viereckende Stab mit seinem
Aufgeklebeten Masstabe ist, c aber stellet den meingen Schieber vor, der durch den
hinter der Machine befindlichen Faden in die Hhe und hernieder gestellt werden kan,
angesehen derselbe oben und unten ber kleine Rollen d, e gehet, und mit seinen beyden
Enden an einander fest geknpfet ist. Damit aber besagter Schieber allezeit fest auf dem
Masstabe wieder gedruckt werde, so sind 2 eiserne Federn hinten [251] her an demselben
gemacht, die in einer Nuthe so lngst des holzernen Stabes eingehobelt worden, auf und
ab gerutschet werden knnen. Oben her bey f ist ein anderer unbeweglicher Arm mit 2
Lcher, wodurch der seidene Faden gezogen wird, feste eingeschraubet, ber demselben
aber befindet sich ein Knopf g, der in einem Loch auf der Hirnseide des viereckenden
langen Stabes sich gedrange herumdrehen let, und um welchen der berflige
Seidenfaden gewickelt werden kan, angesehen das Bleygewichte oder der Senckel k, nicht
lnger vor dem Stabe hangen mu als da er juste in seinen Vibrationen bey dem Zirckel
h, welcher unten auf dem Masstabe gezeichnet ist, vorbey streiche, ohne welches die
Vibrationes nicht wichtig seyn wrden und bey deen Vorbeypassierung man jedes Mahl
den Tact schlagen und also die Geschwindigkeit des Tempo erkennen mu. Damit man
aber die eigendliche Einrichtung des beweglichen Schiebers desto beer sehen knne, so
habe sie in nachfolgender Figur [Enlargement of K and C from Figure 2 - BJ] besonders
abgezeichnet.
Wie auch das Bleygewichte nach seiner nathrlichen Gre. Aus dem Masstabe, welcher
auf dem viereckenden Stock lngst herunter stehet, siehet man brigends wie die Mensur
sich immer verkrze, nachdem sie weiter herun[252]ter kommet, wie ich solche nach
eigendliche Verjngung nach angeben eines besondern Maasstabes aufgetragen. In
*

Transcription courtesy of Dr. Paul Peucker, Archivist of the Moravian Church, Northern Province, Bethlehem,
Pennsylvania, USA.

188

TvM_15_#3_nov_2010_7.indd 188

15-11-2010 12:27:57

dutch journal of music theory

den runden Zirckel fr welchem das Bleygewichte sonst zu vibriren pfleget, setzet man
Ziehrats wegen ein klein rund Spiegel Glas, damit man die Vibrationen desto beer sehen
kan. Unter demselben aber stehet nachfolgendes Register von Tntzen, deren Tempo man
zu sehen verlanget, und weil der Raum in der Abzeichnung alhier zu klein geween, so
will sie folgendermaen hier einrcken.
Verlangt man nun dieem nach das Tempo eines Tantzes, e.g. Menuets zu sehen, so
faet man die Schnur so ber beyde Rollen hinten an dem Instrumente gehet und den
untersten beweglichen Schieber anziehet, an, und stellet solchen ber die Zahl wo 48 stehet,
siehet zu, da das Bleygewichte nicht lnger an seinem Faden als vor dem Spiegelgen,
wie auch nicht krtzer hange, giebt demselben einen Sto, oder lt es seine Vibrationes
machen, und schlgt so offt der Senckel bey dem Spiegelgen vorbey fhret den Tact, so
wird das rechte Tempo vor einen solchen Tantz herauskommen, welches die Operisten
so wohl als Musicos in Ordnung und einer gleichen Mensur halten, auch sonsten in der
Music nicht wenig Nutzen kan. Es ist brigends aus denen Gesezen der Bewegung und
der Mechanic bekant, da ein Senckel in seiner Schwenckung nicht mehr Zeit erfodere,
wenn er ein groes Zirckelstck fhret oder wenn er nur ein kleines anweiet, inmaen
er in dem ersten Fall desto geschwinder, im lezten aber desto langsamer gehet, und wenn
anderst eine lnge [253] von Faden, oder ein Centrum oscillationis behalten werden,
einerley Zeit Versaumung erfordert da entwegen darff man also bey dieem Instrument,
so seyn Erfinder Mons. Feuillet, cronometre betittult, nicht frchten, da der Tact ungleich
werde angegeben werden, sintemahl die Schwenckung eben so viel Zeit wenn sie weit
ausgreiffet, oder wenn sie nur ein kleines Zirckelstck abschneidet und einen schwachen
Sto bekommen, oder auch wenn sie in der Lnge allmhlig nachlet, erfodert, und den
Tact immer einerley accurat angiebt bi der Senckel sich gar nicht mehr rhret, das doch
eine ziemliche Zeit whren kan. Es wird brigends die Machine selbst in Paris von dem
Autore verfertigt, woran es gleichfal bekommen, und welcher zu deutlicherm Unterricht
noch nachfolgende Beschreibung gemeiniglich mit bey leget:
[French text explaining the crescents that accompany the pouce numbers in Figure 3.]
[254 bottom] Da man nun also mit dieer Machine die Tnzer und Musicos nicht allein
ber einen Kamm, auch ohne Abrede bringen kan, sondern auch bey allen Concerten die
Strittigkeiten wegen des rechten Tempo vermindern werden, ein solches siehet man nicht
nur gar leichtlich aus der Beschreibung, es dienet aber auch diejenige, so noch nicht gar
feste und richtig im Tacte sind, zu strcken und das Geklopfe bey einer Music berhoben
zu seyn.

189

TvM_15_#3_nov_2010_7.indd 189

15-11-2010 12:27:57