D O C U M E N T 2 1 6 M A R C H 1 9 2 6 2 2 7 happily—an art you do not achieve as perfectly as most others of your sex for, I daresay, your lively head does not like to let itself shut down! Remember the Asi- atic past [3] then you will fully experience the elusiveness of all life forms and be- come well. For now, warm regards, your A. Einstein 216. To George Y. Rainich [Berlin,] 8 March 1926 Dear Colleague, Just now I received your note for Physika.[1] The difficulty in explaining the dif- ference in the physical nature of the two electricities lies very deep. For the diffi- culty comes from this: the electric density scalar determines the sign of the electricity. The electric field determines the electric density vector according to the equation .… (1) From this the density scalar σ is determined by the equation .… (2) The sign of σ remains undetermined. Laws that only contain , and their de- rivatives hence cannot differ for the σ-fields of both signs. Thus, such laws can never explain that an electron of opposite sign and of the same mass does not exist.[2] A theory that explains the electrons must contain equations in which the sign of σ plays a role. If, furthermore, the sign of σ is supposed to be determined by the field, then (2) must be supplemented: Assume a distinction had been made a priori regarding light cone K of point P, a distinction between the backward- cone V and the forward-cone N. In reality, such an a priori distinction has not been necessary until now, but we can es- tablish: σ is positive if the (timelike) current vector ( ) of point P lies in N, but negative if it lies in V. I tried to set up field equations which depend on a a nonsymmetrical function of σ. But I did not succeed.[3] — ∂xν ∂ f μν Iμ = σ2 gμνIμIν = gμν f μν Iμ