So, this piece was actually dedicated to Jean Dreyfus--as in the infamous Dreyfus affaire in France over a man in the french army (jewish ethnically whereas the army was heavily influenced by the French Catholics) who was unfairly accused of treason, and many of the artists got up in arms about it. Still he was sent to a prison camp in South America for many years until he was finally acquitted. (I'm just proud I still remember this)
Also, my mother used to play this as I was growing up. Her score had no indication of greater work or composer, but now I know who wrote it.
It is in composite ternary form. The first A section is a rounded, open binary. The a of A is composed of two symetric, parallel phrases of four measures each which form a phrase group. The first is in G and ends in an authentic cadence on the tonic that in my opinion is weakened by the use of la (e) in the V. The secound four measures modulate to e minor and end on a V in e. The b of A is very developmental and even expository as the development births new turns and arcs of melody, using the basic material of a as its guide. It runs for fifteen measures without real cadence, the melody constantly pushing forward, then ends on a dm chord in 2nd inversion, thus returning to a' with the melody is set an octave higher and not quite back in G major. The first phrase of a' ends with a PAC in a minor and falls into a c minor chord (chromatic mediant?) and then returns to G for the second phrase where it ends in a PAC (with that e still in the V). Though the transition is long, it forms only one real phrase and is tonally unstable. It draws its line from the melody of a and leads back to an altered return of a; therefore, I judge it to be rounded binary.
The B section has no transition leading into it. The a of B is composed of two parallel eight measure phrases in d minor which are practically identical except for the addition of a motive in the upper register of the right hand in the second phrase. Both phrases have sort of an authentic resolution melodically, except that the chord preceding is a VII in dm with a do in the base, plus the bass of the tonic chord at the end of each eight measures is G. We again have a flowing b section that develops on the material of a, using the same rhythm up until the end and transition back in to a'. There's a sort of rest at the end of 13 measures, because it lands on a d minor chord, but it moves on for three measures more and then into two measures of transition for a one phrase return of a where the chords are slightly different, thought it ends with that same odd cadence on the tonic, thus making this a closed (or sectional) rounded binary.
The return of A is about the same, but there is a large extension at the end where the melodic gesture is shortened and repeated for eight measures and then changes texture, fading into real terminative section which lasts about 17 measures--the texture gets very thin and gives a sense of winding down with little returns of familiar activity to end finally with an open fifth on the g in the bass and a major D7 in the right hand--so, sort of returning to original tonic.
The ternary form is closed and composite (since we have these nice binary forms always returning to their original key, thus closing the major sections). I love this piece. When I hear it, I see my mother, I see my youth. I remember the piano when it was something mystical, when music was a strange and uncharted water.