(1) Matter accreting to the inclined dipole, forms nonaxisymmetric structures in the magnetosphere of the star. The flow has different shapes at different density levels. At a low density, matter covers the entire magnetosphere. At larger density levels, matter is seen to accrete in streams. The streams are narrower at higher densities and wider at smaller densities. The streams may obscure the light emitted from the surface of the star or from hot spots on the surface of the star thereby causing variability of the star. 

(2) At a small inclination angle Q = 5°, the densest matter accretes in two streams which precess around the star's rotation axis with angular velocity larger than that of the star. The high velocity of precession is connected with the fact that inner regions of the disk rotate faster than the star, while the streams rotate with intermediate velocity. The lower density matter covers the whole magnetosphere, but forms lower-density windows, which also precess around the rotational axis. Precessing streams and windows may lead to quasi-periodic variability of the star. 

(3). At larger inclination angles, Q= 15° and Q = 30°, the densest matter also forms streams but they settle at a particular location, some 20° - 30° downstream (counter-clockwise) from the (W, m) plane. For such inclination angles, the streams may modulate the emission from the surface the star or hot spots on the star's surface. Thus, in addition to the main period associated with the rotation of a hot spot, an additional, shorter scale period may be observed. If accretion is not stationary, then intervals of quasi-variability may be observed connected with temporary precession of the streams.

(4) At the even larger inclination angle, Q = 45°, matter flows in two or several streams. These streams may also modulate the star's emission. Variability of the accretion rate will lead to variation of the flow pattern in the magnetosphere. Recently Muzerolle, Calvet, & Hartmann (2001), suggested that some observations of CTTSs may be explained if the funnel flows have a multiple stream geometry. 

(5) For Q = 60°, the matter flows to the surface of the star in a funnel flow consisting of two or four streams. A significant fraction of the total accretion flows directly to the nearby pole as a result of the high inclination of the dipole. For Q = 75°, matter flows to the star along funnel flows consisting of two streams. The geometry these streams is different from those at small Q: they are basically located in the plane of the disk. 

(6) The inner regions of the disk are often warped or tilted. This tilt is connected with the fact that matter leaving the disk and entering the magnetosphere has a strong tendency to co-rotate with the magnetosphere, so that the inner tilted disk has an axis close to that of the dipole m axis. We never observed the warp in the direction predicted by theory because the twist of the magnetic field lines, which is necessary for such warp, is always small. 

(7) The interaction of the star's magnetic field with the inner regions of the disk lead to magnetic braking of the disk matter. At relatively small inclination angles, Q = 0° - 30°, this leads a redistribution of the density in the disk with lower density in the inner regions, r < 4, and higher density (the ring) near magnetospheric boundary, at r ~ 1  - 1.5. Such density distribution commonly observed in our two-dimensional simulations for different time-scales, from T = 1 > 1 to T = 50 rotations. At larger inclination angles, Q = 45° - 75°, the density in the region r < 4 is also decreased, however the dense ring of matter does not form near the magnetospheric boundary. Instead, two high density streams are observed in the polar region. 

(8) Magnetic braking leads to significant departure of the angular velocity of the disk from Keplerian in the region r < 2 for smaller inclination angles, and for r < 3 for larger inclination angles. At Q = 60°  and  Q = 75° a significant part of the disk up to r ~ 2.4 almost co-rotates with the star. 

(9) The accretion rate in the established flow  dM /dt is approximately the same for a wide variety of inclination angles, Q = 5° - 45° with slightly larger values for Q  = 60°  and Q  = 75° . We conclude that dependence of the accretion rate on Q is not very high. These values approximately coincide with the accretion rate obtained in analogous two-dimensional simulations. The accretion rate associated with pure hydrodynamic accretion is about 10 times smaller. 

(10) We observed that angular momentum is transported from the disk to the star and that it is carried mainly by the magnetic field near the stellar surface. The same situation was observed in our two-dimensional axisymmetric simulations. For the considered small angular velocity of the star, the angular momentum flux relative to the Z axis gives a positive torque N_z which acts to spin-up of the star for all Q. In addition, we have derived from our simulations the torques N_x and N_y which are found to be non-zero for non-zero inclination angles Q. These torques act to shift the  direction of the angular momentum of the star. 

 Most of simulation runs were done for T = 5 - 12 rotations of the inner radius of the disk. This time is sufficient for understanding the physics of the process, because the initial relaxation occurs in T ~ 0.5 - 2 rotations depending on Q. Two-dimensional axisymmetric simulations, which were done up to T ~ 50 - 80 rotations, have shown that the magnetospheric flow settles at T ~ 1 and later all features of the flow are very similar during many many rotations. This provides support for physical applicability of our simulation results. Nevertheless, one of the planned developments is to generate longer runs with a steady inflow of matter from the disk.