In Part II of these two sequential papers the details of two different planar defects are presented and discussed in view of their relation with the martensitic transformation. It is shown that the lattice displacements and habit planes of the defects can be explained by assuming non-conservative anti-phase boundaries existing in the parent B2 phase, that is, prior to martensitic transformation. The main difference between both defects is the sequence of shuffled (001) planes which remains unchanged in the first type but skips one plane in the second type. The microtwin sequence is also clearly affected by the second type. (C) 1998 Acta Metallurgica Inc.