1 1 4 D O C U M E N T 9 4 O C T O B E R 1 9 2 5 94. To Hendrik A. Lorentz Berlin, 21 October 1925 Dear Mr. Lorentz, It pleased me greatly that you have read that paper and that you attached your own considerations[1] to it.[2] I must admit, though, that unfortunately my confi- dence in the physical significance of the formulas is already badly shaken. If one does not want to presuppose the vanishing of the ’s[3] —as I found only after the paper had been published—one can proceed as follows: If one assumes , then one simply obtains instead of (10a) Instead of (4) one gets . The system of equations thus becomes considerably simplified by introducing the ’s. By limiting oneself to first approximation, (18) can be retained, where instead of (19) one gets Equation (17) is left standing. This way the calculation is surely the simplest.— In the case that the ’s are not assumed to vanish, it is now possible to elimi- nate the ’s. If one assumes[4] which in fact has as the consequence, then according to (1) and (2), ϕα Γμα σ δμ σϕα + Γμα σ * = ∂gμv ∂xα ---------- -– gσxΓμα σ * gμσΓαν σ * + + 0. = 0 ∂Γμν α * ∂xα -------------- -– Γμβ α x*Γαν β x* ∂Γμα α * ∂xv --------------- Γμν α *Γαβ β x*· – + + © ¹ § ∂ϕμ ∂xv --------- ∂ϕv ∂xμ¹ --------· -– © § + = Γμα σ * 1∂2ϕμ- 2 --------------- - v ∂xα 2 ∂ϕμ ∂xv --------- ∂ϕv ∂xμ¹ --------· -– © § + 0. ……(2) = ∂ϕμv ∂xv ----------- 0 = ……(1) ϕμ ϕμv ∂ϕμ ∂xv --------- ∂ϕv ∂xμ -------- -– fμv, = ∂fμ- v ∂xσ --------- ∂fνσ ∂xμ --------- - ∂fσμ ∂xv ---------- + + 0 =