I N T R O D U C T I O N T O V O L U M E 1 5 x l v object and instead introduced both an affine connection and a metric tensor as fun- damental and independent objects on which to base the theory.[8] As in his previous papers on affine field theory, Einstein finished by noting that it would now be nec- essary to investigate whether the theory allows for spherically symmetric solutions that could describe electrically charged particles. In a letter to Besso (Doc. 34), he likewise alluded to the critical question of whether the theory predicts the existence of quanta. For Einstein this meant, first and foremost, that solutions representing electrons have a uniquely determined electric charge and rest mass, and that no continuum of charge and rest mass exists. Alas, as before, Einstein soon lost faith in his latest approach. In letters to Paul Ehrenfest (Doc. 71), Eddington (Doc. 91), and Lorentz (Doc. 94), he abruptly dis- owned the theory writing this to Lorentz must have been particularly galling, as Lorentz had just sent a ten-page letter in which he discussed the theory in meticu- lous detail (Doc. 90). Abandoning yet another stillborn unified field theory renewed internal doubts regarding the very basis of Einstein’s earlier attempts at unification. Instead of im- mediately returning to the fray with a modified approach he now endeavored to give a general argument why all these attempts had to fail. In his contribution to the Festschrift honoring the fiftieth anniversary of Lorentz’s doctorate (Einstein 1925w, [Doc. 92]), Einstein aimed to prove a meta-theorem about a whole class of theories that included, as special cases, Einstein-Maxwell theory, Eddington’s theory, and his own affine field theories, as well as his newest approach based on a mixed metric-affine geometry. The theorem stated that any theory that represents the electromagnetic field by an antisymmetric second rank tensor, and that has a solution representing a particle with the charge and mass of an electron, must also have a solution corresponding to a particle with the same mass but opposite (i.e., positive) electric charge. And since Einstein assumed, as did every physicist at the time, that the only two fundamental particles were the electron and the proton (with the latter about two thousand times heavier than the former), he judged that all these theories were empirically inadequate. Thus again we see that Einstein’s litmus test for any field theory was that it should predict the discrete (quantum) fea- tures of the known elementary particles. Einstein’s Lorentz Festschrift paper led Rainich to write yet again. He had penned a one-page response (Rainich 1926b), and sent it to Einstein on the same day that he sent it for publication to Adriaan Fokker, the editor of the journal in which Einstein’s paper (together with all the others presented on the occasion of Lorentz’s anniversary) were published. Rainich disputed Einstein’s interpretation of his meta-theorem, which brought about the second, even more fruitful, phase of their exchanges during which they discussed the difference between linear and