2003ApJ...586..464B

We propose a simple interpretation of the rotation period data for solar- and late-type stars. The open cluster and Mount Wilson star observations suggest that rotating stars lie primarily on two sequences, initially called I and C. Some stars lie in the intervening gap. These sequences, and the fractional numbers of stars on each sequence, evolve systematically with cluster age, enabling us to construct crude rotational isochrones allowing ``stellar gyrochronology'', a procedure, on improvement, likely to yield ages for individual field stars. The age and color dependences of the sequences allow the identification of the underlying mechanism, which appears to be primarily magnetic. The majority of solar- and late-type stars possess a dominant Sun-like, or interface, magnetic field, which connects the convective envelope to both the radiative interior of the star and the exterior, where winds can drain off angular momentum. These stars spin down Skumanich style. An age-decreasing fraction of young G, K, and M stars, which are rapid rotators, possess only a convective field, which is not only inefficient in depleting angular momentum but also incapable of coupling the surface convection zone to the inner radiative zone, so that only the outer zone is spun down, and on an exponential timescale. These stars do not yet possess large-scale dynamos. The large-scale magnetic field associated with the dynamo, apparently created by the shear between the decoupled radiative and convective zones, (re)couples the convective and radiative zones and drives a star from the convective to the interface sequence through the gap on a timescale that increases as stellar mass decreases. Fully convective stars do not possess such an interface, cannot generate an interface dynamo, and hence can never make such a transition. Helioseismic results for the present-day Sun agree with this scheme, which also explains the rotational bimodality observed by Herbst and collaborators among pre-main-sequence stars and the termination of this bimodality when stars become fully convective. This paradigm also provides a new basis for understanding stellar X-ray and chromospheric activity, light-element abundances, and perhaps other stellar phenomena that depend on rotation.