The magnet did not close. Here is the arithmetic.

DRIFT was originally specified around magnetic vestibular stimulation. Ionic current in the inner-ear endolymph crossing a static magnetic field takes a Lorentz force that deflects the cupula the way real rotation does. The effect is documented and replicated in high-field MRI, where stationary subjects report continuous rotation. Nothing about that is in doubt.

The question is whether the field can be delivered from an aircraft. It cannot. Holding the threshold field of one tesla across a 9.5 metre footprint from four metres needs a dipole moment of 1.197 × 10⁹ ampere metres squared. That is 1,629 clinical MRI magnets.

Three closures, and any one of them ends it. At a credible engineering current density the winding cross-section comes to 7.94 square metres, which is larger than the 0.40 metre radius coil that has to contain it. The geometry fails before the mass does. The field inside the winding would be 9,978 tesla against a destructive record of about 1,200, at nearly twenty thousand times the yield strength of maraging steel. And the stored energy is 5.97 terajoules, which is 1.4 kilotonnes of TNT sitting under a 24 kilogram aircraft.

We checked the pulsed variant too, because it is the obvious next question. Induced-field stimulation depends on dB/dt rather than B, so a fast pulse does reduce the peak field. It reduces it to between 54 and 327 MRI magnets, and it crosses peripheral nerve and muscle thresholds before it achieves any vestibular selectivity. The result would be involuntary muscle contraction and pain, not a balance effect.

There is a fourth closure that holds even if the first three were free. The five gauss contour would reach 78 metres. That is the boundary inside which pacemakers, implanted defibrillators and cochlear implants are unsafe, and inside the one tesla contour loose steel becomes a projectile. A weapon that kills anyone in the crowd with a cardiac device is not a non-lethal effector under any definition. It fails our founding constraint before it fails physics.

So the delivery changed and the target did not. DRIFT still writes to the vestibular organs of the inner ear. It reaches them with air-conducted sound at 500 Hz instead of with a magnetic field, because that is the frequency the utricle and saccule are most sensitive to and because 128 decibels can be flown while one tesla cannot. A contradicting visual field runs against it. The airframe, the doctrine, the footprint and the discrimination story are all unchanged. The full analysis is a runnable script, not a claim: models/drift_magnetics.py.

We publish this because the first question any competent evaluator asks about a vestibular weapon is whether the field closes. We would rather show the arithmetic than be asked for it.