'Sonata Pathetique' (op. 13) we find the first subject in C minor:—
![\new ChoirStaff << \override Score.Rest #'style = #'classical
\new Staff \relative g' { \key c \minor \time 4/4 \partial 4.
g8-. c-. d-. | ees4.^\( f8 d4. ees8\) |
c2 \appoggiatura d8 c b c d | s_"etc." }
\new Staff \relative c { \clef bass \key c \minor
r8 r4 | c8^\( ees g c\) c,^\( f g b\) |
c, ees g c ees g ees d | s } >>](http://upload.wikimedia.org/score/p/h/phjhb6bsdwg8ivpmp458shf8fr4ymfx/phjhb6bs.png)
this is of 17½ bars in length and ends with a full close in the key. Six bars follow, modulating into E♭, where we find the second subject, which is of unusual proportions compared with the first, consisting as it does of three separate themes:—
![\relative g' { \override Score.TimeSignature #'stencil = ##f \time 4/4 \key c \minor \mark \markup \small "(1)"
g4 bes2 bes4 ~ | bes8 c d ees f g aes f | aes g f ees ees d c d |
f ees d c bes4 }](http://upload.wikimedia.org/score/4/u/4uenjuk2iqnaukpyvjesz3tke0mctei/4uenjuk2.png)
etc.
![\relative g' { \override Score.TimeSignature #'stencil = ##f \override Score.Rest #'style = #'classical \time 4/4 \key c \minor \mark \markup \small "(2)"
<< { \override TupletNumber #'stencil = ##f \override TupletBracket #'bracket-visibility = ##f \tuplet 3/2 4 { g8 ees f g f ees } a2 |
\tuplet 3/2 4 { g8 ees' f g f ees } a2 } \\
{ \override TupletNumber #'stencil = ##f \override TupletBracket #'bracket-visibility = ##f ees,,4 r \tuplet 3/2 4 { r8 bes' c d c bes } |
ees4 r \tuplet 3/2 4 { r8 bes' c d c bes } } >> }](http://upload.wikimedia.org/score/3/7/37094ipsiyeihj7f5cmo154ulzwutvp/37094ips.png)
etc.
![\relative b { \override Score.TimeSignature #'stencil = ##f \time 4/4 \key c \minor \mark \markup \small "(3)" \partial 4
<< { bes4 <f' d> q q q | <bes ees,>2. q4 | <aes ees>2 <a f ees> | <bes f d>2. }
{ s4 <bes, aes> q q q | <bes g>2. g4 | c2 ces bes2. } >> }](http://upload.wikimedia.org/score/n/h/nhb8lvbrq2nchivvsm9bj17hr2rh43p/nhb8lvbr.png)
etc.
After this we return to the 1st subject, which ends just as before. A new start is then made with a third subject (or pair of subjects?) in A♭:—
![\relative c' { \override Score.TimeSignature #'stencil = ##f \time 4/4 \partial 4. \key c \minor \clef bass
<< { s4. | c2 f bes, ees | aes, des4 c | bes aes g ees } \\
{ aes,8 c ees | aes2 des, g c, | f4 bes,2 c4 | des d ees aes, } >> }](http://upload.wikimedia.org/score/1/1/11u7fpvnxk3se1fu5h21qbmeoo49gsn/11u7fpvn.png)
this material is worked out for 24 bars and leads to a prolonged passage on a chord of the dominant seventh on G, which heightens the expectation of the return of the 1st subject by delaying it. On its third appearance it is not played quite to the end, but we are skilfully led away, the bass taking the theme, till, in the short space of four bars, we find the whole of the 2nd subject reappearing in C major. Then, as this is somewhat long, the 1st subject comes in again for the fourth time and a Coda formed from the 2nd section of the 2nd subject concludes the Rondo with still another 'positively last appearance' of No. 1.
Beethoven's Rondos will all be found to present but slight modifications of the above form. Sometimes a 'working-out' or development of the 2nd subject will take the place of the 3rd subject, as in the Sonata in E (op. 90), but in every case the principal subject will be presented in its entirety at least three times. But as this was apt to lead to monotony—especially in the case of a long subject like that in the Sonata just quoted—Beethoven introduced the plan of varying the theme slightly on each repetition, or of breaking off in the middle. It is in such delicate and artistic modifications and improvements as these that the true genius shows itself, and not in the complete abandonment of old rules. In the earliest example we can take—the Rondo of the Sonata in A (op. 2, No. 2), the form of the opening arpeggio is altered on every recurrence, while the simple phrase of the third and fourth bars
![\relative g' { \override Score.TimeSignature #'stencil = ##f \override Score.Rest #'style = #'classical \time 4/4 \partial 4 \key a \major
gis4 | a b cis d | \appoggiatura cis8 ees4. d16 cis b4 r }](http://upload.wikimedia.org/score/r/0/r0laa177g63wx3ssatw2pjafj5kpilr/r0laa177.png)
is thus varied:—
![\relative a' { \override Score.TimeSignature #'stencil = ##f \override Score.Rest #'style = #'classical \time 4/4 \key a \major
a8. e16 b'8. e,16 cis'8. a16 d8. a16 | e4. d16 cis b4 r }](http://upload.wikimedia.org/score/c/7/c78psz4cza415b8r6r169inynfzwfcp/c78psz4c.png)
![\relative a' { \override Score.TimeSignature #'stencil = ##f \override Score.Rest #'style = #'classical \time 4/4 \key a \major
a8. cis32 b b8. d32 cis cis8. e32 d d8. dis16 | e4.. d32 cis b4 r }](http://upload.wikimedia.org/score/k/7/k7g122mdsgkf6kpwyzkphor3r8ach13/k7g122md.png)
![\relative a' { \override Score.TimeSignature #'stencil = ##f \override Score.Rest #'style = #'classical \time 4/4 \key a \major
r8 gis16 a r8 ais16 b r8 bis16 cis r8 cis16 d | e4 ~ d16 dis32 e fis e d cis b4 r }](http://upload.wikimedia.org/score/2/s/2s15nw6piwkyyymcpuziio4xsf9lw41/2s15nw6p.png)
![\relative a' { \override Score.TimeSignature #'stencil = ##f \override Score.Rest #'style = #'classical \time 4/4 \key a \major \override TupletNumber #'stencil = ##f \override TupletBracket #'bracket-visibility = ##f
\tuplet 6/4 2 { a16 b a gis a ais b cis b ais b bis cis d cis b cis a d e d cis d dis | e fis e dis e eis } fis e d cis b4 r }](http://upload.wikimedia.org/score/5/g/5g3fpn5wmbee8tctdp31tky8dp67zc4/5g3fpn5w.png)
In the Rondo of the Sonata in E♭ (op. 7) again, we find the main subject cut short on its second appearance, while on its final repetition all sorts of liberties are taken with it; it is played an octave higher than its normal place, a free variation is made on it, and at last we are startled by its being thrust into a distant key—E♮. This last effect has been boldly pilfered by many a composer since—Chopin in the Rondo of his E minor Pianoforte Concerto, for instance. It is needless to multiply examples: Beethoven shows in each successive work how this apparently stiff and rigid form can be invested with infinite variety and interest; he always contradicted the idea (in which too few have followed him) that a Rondo was bound in duty to be an 8-bar subject in 2-4 time, of one unvarying, jaunty, and exasperatingly jocose character. The Rondo of the E♭ Sonata is most touchingly melancholy, so is that to the Sonata in E (op. 90), not to mention many others. There will always remain a certain stiffness in this form, owing to the usual separation of the subject from its surroundings by a full close. When this is dispensed with, the piece is said to be in Rondo-form, but is not called a Rondo (e.g. the last movement of Beethoven's Sonata op. 2, No. 3).
Modern composers, like Chopin, with whom construction was not a strong point, often omit the central section, or third subject, together with the repetition of the first subject which accompanies it, and thus what they call a Rondo is merely a piece on the plan of a French overture; that is to say, having produced all his material in the first half of the piece, the composer repeats the whole unchanged, save that such portions as were in the Dominant are, in the repetition, given in the Tonic. Chopin's 'Rondeau brillante' in E♭, the 'Adieu a Varsovie' indeed all his Rondos show this construction, or rather, want of construction.
[ F. C. ]