Nature of the unusually long solar cycles

The evolution of prolonged and unusually long solar cycles such as 23rd, 20th, and especially 4th is considered. Why the length of the 4th cycle was exceptionally large or really composed of two short cycles? Are the prolonged solar minima can be considered as precursor of low activity of the next cycles? Resolving these puzzles seems to be very important for dynamo theories trying to explain the solar long-term variations. We propose a possible model of the butterfly diagram during unusually long and prolonged cycles, based on (i) the Gnevyshev idea of sunspot distribution over the latitudes, and (ii) the phase differences of the northern and southern hemispheric activities.