The possibility of additional strains in twinned Fe-Ni martensites (32.5 %Ni) has been examined by analysing the traces of the twin interface planes, based on the two surface method with X-ray and optical microscopy techniques, besides electron microscopy of the fine details of twins. It has been verified from the two-surface analysis that the twin interface planes coincide with the (112) twinning plane within a few degrees, and therefore that large deviations between those planes, as detected in a previous electron microscopy, are apparent ones due to some bucklings of foil specimens. On the other hand, electron microscopy suggests the occurrence of slip on a plane including the [111] twinning shear direction, revealing two kinds of twin traces even over a narrow region of the same martensite, and furthermore indicates the existence of planar faults or thin twins on the (121) Or(211)T plane, accompanying streaks normal to the plane in diffraction patterns. From these results, it is suggested that the transformation twins in Fe-Ni martensites are subjected to additional strains by which (Received March 18, 1974) It is well known that transformation twins exist in martensites of many ferrous alloys and steels(1). When one of the present authors found the transformation twins in Fe-Ni martensites, he also detected some deviations of the twin interface planes from the (112) twinning plane, and suggested that the twin plates were subjected to slip along a plane including the [111] twinning shear direction(2). This suggestion was experimentally supported later by Rowlands et al. (3), and they considered the deviation to be an evidence for the existence of multiple lattice invariant strains. Accordingly, phenomenological multiple shear theories of martensitic transformation have newly been developed on the premise of additional strains (4)(5)On the other hand, Patterson and Wayman(6) did not observe any deviation between the twin interface and twinning planes, although they studied on nearly the same composition of Fe-Ni alloy. Recently, Patrician et al. (7) pointed out that such a deviation was due to an experimental error which might be caused by a substantial buckling of foil specimens.In view of the importance of additional strains for the elucidation of the transformation mechanism, it has been desired to check the genuineness of the deviation excluding the effect of specimen buckling. In the present experiment, the interface planes of transformation twins were analyzed by the two-surface method with X-ray and optical microscopy techniques. Consequently, it was clarified that the amount of deviation electron microscopy.The experimental results will be reported below, and discussed on the authenticity of trace analysis by electron microscopy and the multiplicity of lattice invariant strains.An ingot of Fe-32.5 wt %Ni alloy was prepared by melting electrolytic iron and nickel (99.99 %) in a vacuum induction furnace. A part of the ingot was shaped in a plate of 5 mm in thickness by repeating quartz capsules.After...