2 5 2 D O C U M E N T 2 5 6 A U G U S T 1 9 2 8 , , one may thus choose: , , from this as a particular solution , furthermore, , , , and this solution would certainly seem, however, to be acceptable. VI. But your remark that the method of integration suggested indeed has no in- variant character[9] is fundamentally important. I will make an effort to improve that situation. It plays a role especially in the particular solution obtained, where completely unjustified singularities occur, which can then again be can- celed out by the pure gravitation solutions that are to be superposed the latter can likewise now all be given—but I do not yet have a final overview of the whole mechanism, and need to think about it further. Nevertheless, it seemed to me to be not unreasonable to communicate the above to you, so [you can see] that the solubility in principle of the systems obtained is secure. Sending you best wishes from house to house, yours truly, H. Müntz 256. From Chaim Herman Müntz [Berlin-Nikolassee,] 18 August 1928 Dear, esteemed Professor, If for , we go directly to a particular solution,[1] then, apart from the , only the can be considered to be different from 0, and further- more independent of t. In the case of the noncoupled invariant,[2] this leads to U a m 0 = U a 4 2F 4 – U a 4 [ ??]  + = U a 4  2 2 t ----------- - 0 = U 4 m   mm  4  H 4 – = 2F 4 U m 4 –   mm  4  H 4 – 2H 4 U m 4   mm – = = U m 4 0 = H 4 1 3  4  t 3r = =  t ----H 4 m  3 -- x m r x m +   r – log =  2 m t x -------------H - 4 m w/o summ.  3 r x m +  ---------------------- = F 4 t 6 ----r = U 4 m t 3 ----r =  2 m t x -------------Um - 4 w/o summ  3 ------- x m r - = r x m +   –1  a 0  4 j  r = =    h 4 a 