Page:Essays and Addresses.djvu/469

 seems far more probable. He points out—what, indeed, is evident—that there is something frigid, and foreign to the spirit of classical Greek poetry, in the notion of grouping three tragedies under an abstract idea. And he justly remarks that the conjectural reconstruction of theme-trilogies is apt to become a highly arbitrary process. Aeschylus, he thinks, often linked his three plays by fable—as in the Oresteia—but did not invariably do so. Sophocles and Euripides inherited that freedom of choice; with them, probably, the linking of the three plays by story was less frequent than it had been with Aeschylus; this inference is warranted by the extant evidence of their plays and fragments. In cases where the three plays were not linked by fable, we are still at liberty, Günther says, to suppose that the poet chose their subjects with some regard to artistic effects of harmony or contrast. While concurring generally in this view, I think that it requires to be qualified by some further remarks. (1) First, though Welcker's attempt to reconstruct the Aeschylean trilogies, by links of fable or of idea, involves a very large measure of uncertainty—as he himself fully admits,—still he may be said to have proved thus much, that the trilogy in which the plays were linked by fable was the characteristically Aeschylean form of composition. Aeschylus did not always use it; but it was the form distinctively associated with his name. (2) Secondly, the trilogy in which the plays were not linked by fable was characteristically Sophoclean—the form best suited