Physical theories must not introduce as many arbitrary constants as there are phenomena to be accounted for; they must establish connections among the various experimental facts and, above all, must lead to predictions.

In: A. d'Abro, The Rise of the New Physics (Volume One)