IT has been known for some time that the cell membrane of an amœba in water is birefringent. The optical axis lies radially, and the sign is positive with respect to the tangent1 (Fig. 1). I have found that if an Amœba proteus is fixed for about 10 min. in 4 per cent formalin and then immersed in a solution of potassium mercuric iodide of refractive index 1.56, the birefringence of the membrane decreases, and the sign becomes negative (Fig. 2). A similar increase in the negative birefringence of the cell membrane on imbibition with liquids of high refractive index has also been found with red blood cell ghosts2,3, and with sea urchin eggs4. It implies that the membrane has positive form birefringence and negative intrinsic birefringence. I have discussed the implications of this phenomenon, with special reference to the red cell ghost, in another communication3.