Despite the success of the Babcock-Leighton model in explaining some phenomena of the solar cycle, e.g. Hale’s polarity law and Maunder’s butterfly diagram, it has recently encountered a number of observational and theoretical difficulties. The observational difficulties are centred around the different behaviour of bi-polar magnetic regions (BMRs) and uni-polar magnetic regions (UMRs). Babcock explains BMRs as the surface eruptions of toroidal fields wound by the differential rotation and it is clear that in their subsequent evolution and decay they are obviously influenced by this effect. However, UMRs display a sector structure in which the boundaries between regions of opposite polarities appear to be quite unaffected by the differential rotation but closely related to the interplanetary field sectors. Another feature of solar activity unexplained by the Babcock theory is the tendency of sunspots, flares, etc. to reappear at the same latitude during a cycle. Warwick has observed that proton flares show a preference for particular latitudes which may extend over several cycles. Again Bumba et al. have observed that new cycle fields tend to appear in longitude regions where old cycle fields are still visible.