D O C U M E N T 4 5 1 M A R C H 1 9 2 9 4 0 1 451. To Chaim Herman Müntz Saturday [16 March 1929][1] Dear Mr. Müntz, It was indeed a calculation error that caused to appear instead of [2] in . But even in that case, taking the limit of can be carried out in a similar manner. We begin with the equations[3] For the first parenthesis, we find . We note then that from this, we obtain for : , … (2) where is quadratic in the quantities . Furthermore, we require the relation[4] … (3) If we let go to zero in (1), then we find in any case … (1a) It follows from this that the may be expressed as … (4) where  is a scalar density and is a tensor of magnitude 1 that is antisymmetric in all its indices. If we insert (1a) into (3) and replace in these equations with , making use of equations (4), then we obtain 4 equations of second order for , which require us to set , i.e., . We note that in taking this limit, the go to zero as . S /  S     G * G * –  0 = G * G * –    G ** G ** –   + 0. (1) = f  /  –S G * G * 1 2 S /  – Q * + = G  S   D  G *   0 = S /  0 = S   S    x -------- = S    const = S   0 = S   G   – G  