Many minerals of the rhomhie system are of markedly columnar habit, with the axis of elongation, or vertical axis, as one of the biseetrices of the angles between the optic axes. Such a crystal, in a section parallel to the vertical axis, gives straight extinction. It is natural to expect, and is perhaps generally assumed, that a section slightly oblique to the vertical axis will give nearly straight extinction. The object of the present note is to inquire how far this is true, and in what possible cases a serious error may arise.