A cataclysmic variable is a binary star system in which a white dwarf strips material from a companion star. The stolen gas spirals inward, forming an accretion disk that flickers in brightness as the system orbits. Among the brightness patterns, negative superhumps are periodic variations slightly shorter than the orbital period — the disk brightening a little faster than the stars go around.
For decades, the explanation was geometric: the accretion disk tilts out of the orbital plane. A tilted disk precesses — its axis traces a slow cone, like a spinning top. If the precession runs backward relative to the orbit (retrograde nodal precession), it shortens the apparent period, producing the negative superhump. The framework was clean, the mathematics worked, and only one thing was missing: a convincing reason for the disk to stay tilted. Tilt should damp quickly in a viscous accretion disk. The mechanism required a geometry that its own physics would erase.
Vallet, Martin, Lubow, and Lepp propose that the disk is not tilted at all. It is eccentric — elongated, elliptical — and its long axis precesses backward. The distinction matters: a tilted disk wobbles in three dimensions, requiring a force to sustain the wobble. An eccentric disk precesses in the plane, and pressure effects in the gas can drive retrograde apsidal precession even in cool disks without exotic forcing. The geometry that produces the identical observable is the structural opposite of what was assumed.
The consequences diverge sharply. In low mass-ratio systems during outbursts, the disk expands, and different radial regions can precess in opposite directions simultaneously — inner parts going one way, outer parts the other. This allows positive and negative superhumps to coexist temporarily, something the tilted disk model struggles to accommodate. More broadly, negative superhumps appear across diverse cataclysmic variable types, and their prevalence no longer requires every system to maintain a mysterious long-lived tilt.
The general principle: when the assumed geometry is wrong, the inferred mechanism inverts. The same periodic signal, attributed to a wobble in three dimensions, actually arises from a rotation in two. The observable was never ambiguous — the interpretation was. And the error persisted because the tilted disk model was mathematically successful. It predicted the right period. It just required a geometry that erased itself. The eccentric disk predicts the same period from a geometry that sustains itself. The better model is not the one that fits the data but the one whose assumptions survive their own consequences.