We consider the physical implications of the rapid spindown of Soft Gamma Repeater 1900+14 reported by Woods et al .
During an 80 day interval between June 1998 and the large outburst on August 27 1998 , the mean spin-down rate increased by a factor 2.3 , resulting in a positive period offset of \Delta P / P = 1 \times 10 ^ { -4 } .
A radiation-hydrodynamical outflow associated with the August 27th event could impart the required torque , but only if the dipole magnetic field is stronger than \sim 10 ^ { 14 } G and the outflow lasts longer and/or is more energetic than the observed X-ray flare .
A positive period increment is also a natural consequence of a gradual , plastic deformation of the neutron star crust by an intense magnetic field , which forces the neutron superfluid to rotate more slowly than the crust .
Sudden unpinning of the neutron vortex lines during the August 27th event would then induce a glitch opposite in sign to those observed in young pulsars , but of a much larger magnitude as a result of the slower rotation .

The change in the persistent X-ray lightcurve following the August 27 event is ascribed to continued particle heating in the active region of that outburst .
The enhanced X-ray output can be powered by a steady current flowing through the magnetosphere , induced by the twisting motion of the crust .
The long term rate of spindown appears to be accelerated with respect to a simple magnetic dipole torque .
Accelerated spindown of a seismically-active magnetar will occur when its persistent output of Alfvén waves and particles exceeds its spindown luminosity .
We suggest that SGRs experience some episodes of relative inactivity , with diminished \dot { P } , and that such inactive magnetars are observed as Anomalous X-ray Pulsars ( AXPs ) .
The reappearence of persistent X-ray emission from SGR 1900+14 within one day of the August 27 event provides strong evidence that the persistent emission is not powered by accretion .