The Beethoven was glorious, although Beethoven's devotion to that skeleton of a tune has always puzzled me. He used it in four different works: his early Contradanse No. 7, WoO 14 (a collection without an opus number); his second ballet, The Creatures of Prometheus, Op. 43; Sunday's piano variations, Op. 35; and finally for the finale of his “Eroica” Symphony, Op. 55. Each fed on its predecessor, each extending ideas found in its older siblings.

Of course, while the raw theme is as pale as paper, that's precisely its merit for variations. Decorative and alterational possibilities are endless. Complicated themes make for a compositional straitjacket. One has only to consider those full folk melodies used in so many 19th century pieces where there's nothing much to do with them besides repeating them endlessly in different instrumental combinations.

Beethoven covered a lot of emotional ground in his 15 variations. Some are meltingly lyrical, others rambunctious in their extreme flights of virtuoso bravura. They were doubtless meant for his own use, to impress the Viennese and perhaps affright competing pianists. Ursuleasa's mastery of the 25-minute piece was total. Her playing was clean and resonant, free of pounding, even through the bravura passages, and right on the button for tempo selections. It's the kind of Beethoven I have not heard since the heydays of Claudio Arrau and super-musician Clara Haskil herself.

Turning on a dime, Ursuleasa then altered her style of playing for the small Shostakovich cycle, a thing whose demands are totally different from those of Beethoven's pristine heroics. The Preludes date from 1932, when the composer was still in official favor. So among them there's much of the saucy quality of the First Symphony and the early ballet scores such as The Age of Gold. Some are of course more serious than others, and even touched by outright Romanticism, à la Chopin here and there. In that sense, they represent a wonderful grab bag of goodies, and for me his finest work for solo piano.

The surprise is that Ursuleasa managed to slip in and out of the composer's little fancies with so complete a vocabulary of sound. Yet even her staccato playing managed to avoid the usual hammering of the keys, with lots of appreciation for the madcap pieces like the sarcastic nose-thumbing Prelude No. 24. It was all throughly excellent.

Look through a list of Scriabin's solo piano music, and you'll find a blizzard of small pieces, many of them extremely miniature: 199 of them, in fact. Many carry abstract titles, particularly "Prelude" or "Etude." The Three Etudes of Op. 65 were the last of his 26 etudes, clearly intended for performance as a single set. They form a sort of mini-sonata: a long first one (long for Scriabin, that is, at four and a half minutes), followed by two more typically short ones half the length of the first. Like all the very late Scriabin piano music, they are fascinatingly weird, and not a little eerie. One can pick up anticipations of both Varèse and Messiaen among them, although Scriabin is rarely mentioned as a major avant-garde figure of the pre-World War I period. (He died at age 43 in 1915.)

Playing late Scriabin requires a total command of keyboard dexterity and the deepest feeling for dynamics plus great inner appreciation for their quasi-religious and quasi-erotic palette. Ursuleasa accomplished all that with just the right touches of wildness, playing like some dangerous little beast that had crawled out of the bushes at you. As a fan of Scriabin since my teens, it would be hard to imagine a more polished presentation.

Considering all the virtues displayed during the Beethoven, Shostakovich, and Scriabin performances, I was considerably disappointed by Ursuleasa's Brahms. The early virtuoso Brahms sonatas and the big variation sets are terribly difficult to play. Still, if you can manage the notes, they are relatively straightforward. But the 20 short pieces published as his Opp. 116-119 present an entirely different sort of challenge.

Moody in even the fast pieces, umber throughout, they are difficult works to pin down interpretively. (Brahms referred to them as "the catalog of my pain.") Many of the most famous pianists ignore them, or else fail in the attempt to find their way. There exist, for instance, few fully satisfying recordings of those last pieces.

The toxins for Brahms performances are speed and sentimentality. Above all, the music requires a sense of lonely dignity and breathing space enough for the audience to grasp their rich complexities. There are so many shifts in harmony and rhythmic counterpoint that if hurried, they turn into mush. A Brahms Allegro is not as fast as, for example, a Haydn or Beethoven Allegro. After all, how fast is "fast"? These aren't absolute terms. The situation is aggravated by the fact that Brahms did not provide metronome markings.

The other extreme presents another pitfall: wallowing in rubato liberties. Played like languid Rachmaninov, the Brahms pieces become as appetizing as a cold fallen soufflé. Finding one's way through these dangers is horribly tough, and Ursuleasa did not quite succeed in doing so. She played the piano brilliantly, but it just wasn't good Brahms. She sounded harried in the three capriccios and doodled her way through the four intermezzos. Considering her glorious playing of the other three composers, the recital as a whole was definitely of a high order. But I found the Brahms seriously disappointing.