noticing that the Iphigenia in Aulis, like the Rhesus and like no other Greek tragedy, has two alternative openings, one a dull prologue and one a lyrical scene in anapaests under the stars. The general style of the two plays is utterly different; the Iphigenia is most typically late Euripidean; but one would not be surprised to learn that they had both passed at some time through the same revising hand.

This hypothesis seems to work well. But one difficulty remains.

We have so far gone on the supposition that Euripides at twenty-five or thirty perhaps wrote very differently from Euripides at forty-six, and that the manner we call Euripidean is only the manner of his later life. There is nothing improbable in this suggestion, but have we any evidence? Yes, a very little, and unfortunately it does not say what we want. We have some fragments—twenty lines altogether—preserved from the Peliades, with which Euripides won his first victory in the year 455, seventeen years before the Alcestis, and as far as they go they are just in his ordinary manner—a good deal more so, in fact, than much of the Alcestis is. Let us face this difficulty.

The ordinary style of Euripides is full, flexible, lucid, antithetic, studiously simple in vocabulary and charged with philosophic reflection. If we look in his extant remains for any trace of a style, like that of the Rhesus, which is comparatively terse, rich, romantic, not shrinking from rare words and strong colour and generally untinged by philosophy, we shall find the nearest approach to it in the Cyclops. Next to the Cyclops I am not sure what play would come, but the Alcestis would not be far off. It has especially several Epic forms which cannot be paralleled in tragedy. Now the conjunction of these two plays with the Rhesus is significant. The three seem to be three earliest of the extant plays; they are also—if we count the Heraclidae as mutilated—the three shortest. But, what is more important, the Cyclops is not a tragedy but a satyr-play, and the Alcestis is a tragedy of a special sort, written to take the place of a satyr-play. It is a tragedy with some half grotesque figures and a fantastic atmosphere.

This is no place for a close analysis of the diction of the various works of Euripides; but taking one rough test, just for what it is worth, we may try to count the number of words in each play which are not found elsewhere in Euripides. The Medea, a central sort of play, has in its 1419 lines 103 such words. The Alcestis, with 1163 lines, has 122; the Rhesus, with less than 1000 lines, has 177; the Cyclops, with only 701 lines, has actually 220. This calculation is doubtless slightly inexact: in any case it is worth very little until it is carefully analysed. But on the whole it accords with my general impression that the Rhesus in its variation from the Euripidean norm goes further than the Alcestis, and not so far as the Cyclops, and goes in very much the same direction. I feel in the Rhesus a good deal of that curious atmosphere, not exactly comic, but wild and extravagant, which the Greeks felt to be suited to the Satyr horde; the atmosphere normally breathed by the one-eyed Giant of the cavern on volcanic Aetna, or the drunken and garlanded Heracles who wrestles with Death and cracks his ribs for him at midnight among the tombs. The whole scene and setting of the Rhesus; the man-wolf crawling away into the darkness and his two enemies presently crawling in out of the same darkness with his bloody spoils; the divine Thracian king with his round targe that shines by night and his horses whiter than the snow; the panic of the watch, the vaunting of the doomed chieftain, the goddess disguised as another goddess, the thrilling half-farcical scene where the