IN AN INTERVIEW WITH GLENN GOULD ON THE OCCASION OF THE pianist's second recording of Bach's Goldberg Variations, in 1981, Tim Page comments that when he first heard the new performance he was struck by its slow pace, as compared with Gould's 1955 interpretation. Page confesses that he timed the recording and discovered that though the 1981 recording is thirteen minutes longer than the 1955 one, the many reprises Gould observes in this performance but omits in the first-some thirteen da capos and dal segnos-account for much of the difference. Subtract repeats from the fifty- one minutes and fourteen seconds of the second performance, and the two recordings are nearly the same length. Nonetheless, Gould's 1981 rendering seems much slower. "Frankly I can't figure it out ... ," Page concludes his spectacular question. Gould's response-no less impressive, with precise calculations and demonstrative singing-comes down to this: The second recording implements a technique of rhythmic continuity, which gives one constant rhythmic reference point to the entire performance. The variations' theme-indeed played significantly slower in 1981-sets a pulse rate for the rest of the piece, retained in subsequent variations by divisions and multiplications that make subsidiary beats. troughout these varying tempi the feel of the basic pulse rate persists, fostering an experience of elongation, even when the actual pace is quite fast. But in conclusion Gould emphasizes that "[d] escribing a process in this way it sounds so ruthlessly sterile, so relentlessly clinical, and anti- musical. . . . A technique, the idea of rhythmic continuity, is really only useful if everybody does feel it in their bones; experiences it subliminally, in other words."