1998A&A...333..603S

I argue that the rotation of white dwarfs is not a remnant of the angular momentum of their main sequence progenitors but a result of the mass loss process on the AGB. Weak magnetic fields, if present in stellar interiors, are likely to maintain approximately uniform rotation in stars, both on the main sequence and on the giant branches. The nearly uniform rotation of the core of the Sun is evidence for the existence of such fields. Exactly axisymmetric mass loss on the AGB from uniformly rotating stars would lead lead to white dwarfs with very long rotation periods (>10yr). Small random non-axisymmetries (∼10^–3) in the mass loss process, on the other hand, add sufficient angular momentum to explain the observed rotation periods around one day. The process illustrated with a computation of the probability distribution of the rotation periods under the combined influence of random forcing by weak nonaxisymmetries and angular momentum loss in the AGB superwind. Such asymmetries can in principle be observed by proper motion studies of the clumps in interferometric images of SiO maser emission.