Not very often we can see magnetic hysteresis loops of type-II superconductors in their entirety. The published figures mostly present only their upper and lower branch, but their "ends" are usually outside the figure. Yet, in situation when the exciting field amplitude falls between Hc1 and Hc2, the discontinuity (jump) is observed in sample magnetization, even its sign is changed. Anything like this never happens in ordinary magnetic materials: here the two branches are always coinciding at the return point, i.e. at the maximum of the external field. What could be the possible reason for such an amazing behavior?