In this picture (some features of which are included in Fig. 9), the central star is surrounded by a conducting disk with an inner hole (with a radius RX determined by the magnetic field at the inner disk edge and the mass accretion rate through the outer disk). Shielding currents prevent the threading of the disk by field lines, thus an initially dipolar stellar magnetic field (connecting the polar regions by field lines crossing the equatorial plane at a large distance from the star) has to squeeze through the disks inner hole and is strongly compressed in the equatorial plane. In case of a nonideal disk, with some magnetic diffusivity and in the presence of accretion, the field will penetrate the innermost ring of the disk. This field threaded ring is termed the X-region. Since the disk material close to the star will be well ionized and coupled to any magnetic field, the (entire) ring of the disk threaded by the field (in steady state) has to corotate with the star in order to prevent a winding up of the field lines (i.e., the star has to adjust to the angular velocity of the inner disk edge: * = X = (GM*/R3X)½). The radial extent of the part of the disk which is threaded by the magnetic field is of the order of the thickness of the disk.
Material in the innermost part of the X-region rotates at sub-Keplerian velocities and is thus ready to move further in. The magnetic field (which is similar to the undisturbed dipole very close to the star) channels this material in an accretion funnel flow towards some region close to the stellar pole. As the gas moves in, it would like to spin up due to angular momentum conservation. It is, however, attached to the rigidly rotating field lines and thus exerts a forward torque on the star and, more important, on the disk. The angular momentum of the accreting gas is thus stored in the X-region of the disk, which would thus be spun up. At the same time, the field lines threading the outer part of the X-region are inclined to the disk plane by only a very small angle (they have been squeezed through the disk in the equatorial plane from large distances). This part of the disk, rotating at super-Keplerian velocity, can thus launch a magnetocentrifugally driven disk wind: the X-wind. It is powerful enough to open the initially closed stellar field lines (which trace the weak field of the outermost parts of the stellar dipole), allowing the wind to expand. The X-wind efficiently removes angular momentum from the X-region which has been deposited there by the accretion flow.
The density as well as the velocity of the X-wind increase strongly but smoothly towards the polar axis: the X-wind has a core-envelope structure. The degree of concentration towards the polar axis (i.e., the collimation of the flow) increases logarithmically slow with distance from the star. In the X-wind picture, the well collimated jets seen as Herbig-Haro or infrared jets are only the densest axial parts of a more extensive structure. The lower density, slower envelope might explain often observed wide-angle winds (e.g. Kwan & Tademaru 1988) and is supposed to be responsible for the widening of molecular outflow lobes (which are driven by the entire wind/jet). The X-wind driven molecular outflow may thus be regarded as a hybrid of a jet driven outflow and a cavity swept out by a wide angle wind.
It is not yet clear whether the X-wind model really describes the processes at work in a protostellar outflow driving source. Its strength is that it is able to account for many observations in one, fairly self-consistent model (optical observations of time variable accretion/wind phenomena in T Tauri stars, the slow rotation rates of T Tauri stars, a number of the features of jets and molecular outflows, protostellar X-ray activity).