ON THE

THEORY OF THE DIVISION OF THE OCTAVE.







The Theory of the Division of the Octave, and the Practical
     Treatment of the Musical Systems thus obtained. Revised
     Version of a Paper entitled 'On Just Intonation in Music ;
     with a description of a new Instrument for the easy control of
     Systems of Tuning other than the equal Temperament of 12
     Divisions in the Octave. By R. H. M. BOSANQUET, Fellow of
     St. John's College, Oxford.

Contents

PagePage
Mode of expressing intervals in E. T.     ment in regular systems.Positive
     semitones....................................390     and negative systems ..........401
Definitions........................................ 391Application of principle of symme-
Intervals formed by fifths .................. 392     trical arrangement to a "gene-
Regular systems. Theorems α, β ....... 392     ralized key-board" for regular
Regular cyclical systems. Th. i, ii, iii, 393     systems.................... 403
     Th. iv.......................................... 394Application of the positive system of
Multiple systems. Th, v..................... 395     perfect thirds to the "generalized
Formation of major thirds in positive     key-board" (Helmholtz's system,
     and negative systems .................. 396     just intonation)......... 403
Notation for positive regular systems 396Application of the notation of posi-
Notation applicable to all regular      tive systems to the system of 53... 404
     systems, negative as well as positive 397Application of the system of 53 to
Formation of harmonic sevenths in      the "generalized key-board " ...... 405
     positive and negative systems...... 397Application of the system of 118 to
Concords of regular and regular cy-     the "generalized key-board" ...... 405
     �clical systems.............................. 398Application of the negative system of
In regular cyclical systems, to find     perfect thirds (mean-tone system)
     the number of units in any interval     to the " generalized key-board"... 406
     in the scale ................................. 399Application of the negative system of
Employment of positive systems in      31 to the "generalized key-board"406
     music. Rules for thirds ................ 400The investigation of cycles of the
Use of the notation with musical     higher orders-the new cycle of
     symbols...................................... 400     643 and others ......................... 406
Principle of symmetrical arrange-

The mode of expressing Intervals.

     In the original paper presented by the writer to the Royal Society, logarithms were employed as the measure of intervals, as they have been commonly employed by others. Great advantages have been found, however, to result from the adoption of the equal temperament (E. T.) semitone, which is 1/12 of an octave, as the unit of interval. It is the unit most familiar to musicians, and has been found to admit of the expres�sion of the theory of cyclical systems by means of formulae of the simplest character. The writer therefore devised the following rules for the transformation of ratios into E.T. semitones and vice versa, and subse�quently found that De Morgan had given rules for the same purpose which are substantially the same (Camb. Phil. Trans. vol. x. p. 129). The rules obviously depend on the form of log 2. The form of the first rule affords a little more accuracy than De Morgan's.

Rule I. To find the equivalent of a given vibrations-ratio in E. T.
     semitones.

     Take log (ratio), subtract 1/300, and call this the first improved value. From log (ratio) subtract 1/300 of the first improved value and 1/10000 of the first improved value. Multiply the remainder by 40. We can rely on five places in the result.

     The following data are introduced here; they can be verified by numbers given in Woolhouse's tract: -

Fifth = 7.019,500,908,654.
Third = 4-.136,862,861,351.

Five places are ordinarily sufficient.

Rule II. To find the vibrations-ratio of an interval given in E. T.
     semitones.

     To the given number add 1/300 and 1/10000 of itself. Divide by 40. The result is the logarithm of the ratio required. We can rely on five places in the result, or on six, if six are taken.
     Ex. The E. T. third is 4 semitones. The vibrations ratio found as above is 1.259921.

Hence the vibrations-ratio of the E. T. third to the perfect third is very nearly 126:125.

Definitions.

     Regular systems are such that all their notes can be arranged in a con�tinuous series of equal fifths.

     Regular cyclical systems are not only regular, but return into the same pitch after a certain number of fifths. Every such system divides the octave into a certain number of equal intervals.

     Error is deviation from a perfect interval.

     Departure is deviation from an E. T. interval.

     Intervals taken upwards are called positive, taken downwards, negative.

     Systems are said to be of the rth order, positive or negative, when the departure of 12 fifths is � r units of the system.

Intervals formed by Fifths.

     When successions of fifths are spoken of, it is intended that octaves be disregarded. If the result of a number of fifths is expressed in E. T. semitones, any multiples of 12 (octaves) are cast out. Representing the fifth of any system by 7 + δ, where δ is the departure of one fifth expressed In E. T. semitones, we form the following intervals amongst others:-

Departure of 12 fifths = 12δ
     (12 x (7 + δ) = 84 + 12δ, and 84 is cast out).
Two-fifths tone =2 + 2δ
     (2 x (7 + δ) = 14 + 2δ, and 12 is cast out).
Seven-fifths semitone, formed by seven fifths up, = 1 + 7δ
     (7 x (7 + δ) = 49 + 7δ, and 48 is cast out).
Five-fifths semitone, formed by five-fifths down, = 1 - 5δ
     (5x - (7 + δ) = -(35 + 5δ), and 36 is added).

     The seven-fifths semitone will be denoted by s (=1 + 7δ) ; the five-fifth semitone by f (= 1 - 5δ).

Regular Systems.

     The importance of regular systems arises from the symmetry of the scales which they form.

     Theorem α. In any regular system five seven-fifths semitones + seven five-fifths semitones make an exact octave, or 5s + 7f = 12.
     For the departures (from E. T.) of the 5 seven-fifths semitones are due to 35 fifths up, and those of the 7 five-fifths semitones to 35 fifths down, leaving 12 E. T. semitones, which form an exact octave, or,

5(1 + 7δ) + 7(1 - 5δ) = 12.

     Theorem β. In any regular system the difference between the seven-fifths semitone and the five-fifths semitone is the departure of 12 fifths, having regard to sign; or,

s - f = departure of 12 fifths.

     Let δ be the departure of each fifth of the system, then s = 1 + 7δ, f = 1 - 5δ ; whence s - f = 12δ.


Regular Cyclical Systems.

     The importance of regular cyclical systems arises from the infinite freedom of modulation in every direction which is possible in such systems when properly arranged; whereas in non-cyclical systems required modulations are liable to be impossible, owing to the demand for notes lying outside the material provided.

     Theorem i. In a regular cyclical system of the � rth order the difference between the seven-fifths semitone and five-fifths semitone is � r units of the system, or s - f = � r units.

     Recalling the definition of rth order (12δ = � r units), the proposition follows from Th. β.

     Cor.This proposition, taken with Th. a, enables as to ascertain the number of divisions in the octave in systems of any order, by introducing the consideration that each semitone must consist of an integral number of units. The principal known systems are here enumerated:-

Primary (1st order) Positive.
7-fifths semitone
= x units.
5-fifths semitone
= y units
Number of units in octave
(Th. α) 5x + 7y = n
2..........1..........17
3..........2..........29
4..........3..........41
5..........4..........53
6..........5..........65
Secondary (2nd order) Positive.
11..........9..........118
Primary Negative
1..........2..........19
2..........3..........31
Secondary Negative
3..........5..........50

     Theorem ii. In any regular cyclical system, if the octave be divided into n equal intervals, and r be the order of the system, the departure of each fifth of the system is r/n E. T. semitones.

     For departure of 12 fifths = 12δ = r units by definition and the unit = 12/n E. T. semitones;

[therefore] δ = r/n

     Theorem iii. If in a system of the rth order, the octave be divided into n equal intervals, r + 7n is a multiple of 12, and (r + 7n)/12 is the number of units in the fifth of the system.

     Let φ be the number of units in the fifth.

     Then φ(12/n) = 7 + δ = 7 + r/n;

[therefore] φ = (7n + r)/12

and φ is an integer by hypothesis; whence the proposition.

     Cor. From this proposition we can deduce corresponding values of n and r. It is useful in the investigation of systems of the higher orders. Casting out multiples of 12, where necessary, from n and r, we have the following relations between the remainders:-

Remainder of
n....1234567891011
r....5103816114927

     Theorem iv. If a system divide the octave into n equal intervals, the total departure of all the n fifths of the system = r E. T. semitones, where r is the order of the system.

     For by Th. ii. δ = r/n whence

nd = r,

or the departure of n fifths = r semitones.

     This gives rise to a curious mode of deriving the different systems.

     Suppose the notes of an E. T. series arranged in order of fifths, and proceeding onwards indefinitely, thus:-

cgdaebf#c#g#d#a#fcg. . .,

and so on. Let a regular system of fifths start from c. If they are positive, then at each step the pitch rises further from E. T. It can only return to c by sharpening an E. T. note.

     Suppose that b is sharpened one E. T. semitone, so as to become c; then the return may be effected
at the first; b in 5 fifths,
at the second b in 17 fifths,
at the third b in 29 fifths; and so on.

Thus we obtain the primary positive systems. Secondary positive systems may be got by sharpening bb 2 semitones ; and so on.

     If the fifths are negative, the return may be effected by depressing c# a semitone in 7, 19, 31... fifths ; we thus obtain the primary nega�tive systems ; or by depressing d two semitones, by which we get the secondary negative systems ; and so on.

     An instructive illustration may be made as follows; it requires too large dimensions for convenient reproduction here: -

     Set off on the axis of abscissa the equal temperament series in order of fifths, as above, taking about 10 complete periods. If the distances of the single terms are made 1 centimeter, this will take 1.20 m in length, starting from the origin on the left.

     Select a unit for the E.T. semitone of departure, say 1 decimeter.
     Rule a series of lines parallel to the axis of abscissa, at distances representing integral numbers of E.T. semitones, both above and below.

     Rule, parallel to the axis of ordinates, straight lines through the points representing the E.T. notes.

     Enter on the intersections the names of the E T. notes they represent. Thus the notes on the positive ordinate of c are c c# d..., and so on, each pair separated by 1 decimeter, and the notes on the negative ordinate of c are c-b-bb....

     If we then join the c on the left hand of the axis of abscissa to all the other c's on the figure, except, of course, those on the axis, we obtain a complete graphic representation of all the systems whose orders are included. The rth order is represented by lines drawn to the c's in the rth line above, the - rth by the lines drawn to the c's in the rth line below. This illustration brings specially into prominence the singularity of multiple systems, as all the multiples of any system lie on the same straight line with it, and the representation fails to give all the notes of such systems.

Multiple Systems.

     Multiple systems are such that the number of divisions in the octave (kn) in any such system is a multiple (k) of the number of divisions (n) of some other system.

     Multiple systems have not been as yet practically applied.

     These systems are not strictly regular; for though their fifths are all equal, yet they do not form one continuous series, but several. They are strictly cyclical, i. e. they divide the octave into n equal intervals.

     Theorem v. A multiple system, kn, may be regarded as being of order kr, where n is a system of order r.

     For, n being a system of order r, r + 7n is a multiple of 12; [therefore] also k(r + 7n) is a multiple of 12, which is the condition that the system kn be of order kr.

     This is useful in the investigation of systems of the higher orders.

     If n is a multiple of 12, the system is a multiple of the E. T., and of order zero.

     In the illustration described under Th. iv. the notes of a multiple system (kn) are the same as those of system n, until the latter is com�plete. The rest of the representation consists simply of the same notes repeated over and over again. To obtain the rest of the notes we should have to change the starting-point.

     On the whole, we may regard the system kn as consisting of k different systems n, having starting-points distant from each other by 1/k of the unit of the system n.

     It follows immediately that the system kn is of the krth order ; for in every unit of the system n there are k units of system kn ; and so in r units of system n then are kr units of system kn.

     Any system, when n is not a prime, can be regarded as a multiple system.

     Thus the system of 59 is of the 7th order; 118 consequently a multiple system of the 14th order, in which point of view it is of no interest; but, casting out the 12 from the order, it may be also regarded as an independent system of the 2nd order, in which point of view it is of considerable interest.

Formation of Major Thirds in Positive and Negative Systems.

     The departure of the perfect third is -.13686. Hence negative systems (where the fifth is 7 - δ) form their thirds in accordance with the ordinary notation of music. For if we take 4 negative fifths up, we have a third with negative departure (-4δ) which can approximately represent the departure of the perfect third. Thus c# is either the third to a, or four fifths up from a, in accordance with the usage of musicians.

     Positive systems form their thirds by 8 fifths down; for their fifths are of the form (7 + δ), and 8 fifths down give the negative departure (-8δ). Thus the third of a should be db, which is inconsistent with musical usage. Hence positive systems require a separate notation. Helmholtz proposed a notation for this purpose, which, however, is unsuitable for use with written music. The following notation is here adopted for positive systems in general; it is not intended to be limited to any one system, like Helmholtz's. In fact it may, on occasions, be used even for negative systems.

Notation for Positive Regular Systems.

     The notes are arranged in series, each containing 12 fifths, from f# up to b. These may be called duodenes, adopting a term introduced by Mr. Ellis. The duodene

f# - c# - g# - d# - a# - f - c - g - d - a - e - b,

which contains the standard c, is called the unmarked duodene. No distinction is made in these series between such notes as c# and db. These signs refer only to the E.T. note from which the note in question is derived; the place in the series of fifths is determined by the notation. Continuing the series to the right, each note of the next 12 fifths is affected with the mark / (mark of elevation), drawn upwards in the direction of writing. These notes join on to the unmarked duodene as follows: -

e - b - /f# - /c# - /g# . . . ,

and so on.

     Thus /c is 12 fifths to the right of c, and the interval /c - c is the departure of 12 fifths. The next duodene to the right is affected with the mark //, which joins on to the last as before: -

/e - /b - //f# ....,

and so on.

     Proceeding in the same way, we have notes affected with such marks as ///, ////.

     Return to the unmarked duodene, and let it be continued to the left; the notes in the next duodene on the left are affected with the mark \, (mark of depression), drawn downwards in the direction of writing. The junction with the unmarked duodene will be

\c - \g - \d - \a - \e - \b - f# - c# ....

     The next junction on the left will be

\\e - \\b - \f# . . ;

and, proceeding in the same way, we have notes affected with such marks as \\\, \\\\.

     Thus c - e\ is a major third determined by eight fifths down in the whole series; and \e will have the departure (- 8δ) from the E.T. note e derived from c.

Notation applicable to all Regular Systems, Negative as well as
Positive.

     As this notation simply consists of a determination of position in a continuous series of fifths, it may be applied to all regular systems, positive or negative ; but, it is not commonly needed for negative systems, it is not generally applied to them.

Formation of Harmonic Seventh in Positive and Negative Systems.

     The harmonic seventh is the interval whose ratio is 7:4. It affords a smooth combination, free from beats.

     The departure of the harmonic seventh from the note which gives the E. T. minor seventh is -.31174 (Rule I.).

Concords of regular and regular cyclical systems....398.

site index

Hosted by www.Geocities.ws

1