Foote and Spalding, Modern Harmony (1905) (2) note 1

In the previous post on Foote and Spalding's repertory examples of dominant ninth chords, I included this figure from their page 195:

This excerpt is certainly interesting, especially for its ascending move from the ninth, but Schubert's subsequent treatment of it in this variation movement from the A-minor sonata, D. 845 suggests that he was not thinking of it as a dominant ninth chord—or else we might say that he could have been in sympathy with those theorists who lumped the dominant ninth chord in with the dominant seventh and leading-tone seventh chords, and maybe even the ii chord (that is, any and all tertian entities that can be extracted from the dominant ninth chord).

Here is the context, the opening phrase of the theme's B-section, in which the ninth is part of a short "standing on the dominant" figure (the term is Caplin's; such figures to begin the B-section of a small binary or ternary form were stereotypical).

Already in the first variation, the bass changes the harmony to ii and thus A4 to the consonant fifth.

Variation 2: As in variation 1, but now the dissonance is gone!

Variation 3: G-F against Ab is brought very much into the foreground. The G is so obviously heard as an accented dissonant neighbor that the underlying chord in the first bar must be heard as vii°7.

Variation 4: Here the expressive idea of ascent takes over, but without the simple ^6-^7 figure of the theme. We can certainly hear the first bar as Ab: V9, with indirect resolution of the ninth in both fifth and sixth octaves in bar 2. Schubert continues to make the expressive point with the long scale that closes the phrase and leads to Eb7.

Variation 5: A predecessor to Schoenberg's fourth inversion of the ninth chord in Erklärte Nacht? No, a 2-3 bass suspension, A2-G2 against B3. And of course ^6 goes down, not up as in the theme.

Variation 5, end: As it would seem only Schubert can do, the end of the final variation leading into the coda turns unexpectedly magical. Not at (a), where he changes the mode to minor—but he does introduce a rising-scale figure that reaches Ab5 (note that the accented gesture is G and Ab). There is a source in the theme: the harmony is D minor with A4 at the top at the equivalent point that he reaches Db in this variation, but Schubert makes much more of it here than in the theme or earlier variations. At (b) we reach G5-Ab5 again, but now also get the reverse over dominant harmony. Then suddenly it is off into the sixth octave, A-natural, and a transcendent IV (no need for the triad's fifth; what we hear is without any doubt). Finally, at (d) even higher, but then with firm and continuing, if quiet,  descent into the final tonic.

Coda: Schubert isn't done yet. At (e), the old-fashioned turn to the subdominant in the prolongation of the structural tonic brings ^5-^6-^5 in F and so a little coloration of V9 in that key. At (f), Schubert reminds us of that transcendent IV from (c), again at (g), and even into the final seconds at (h).