the story as the widely differing versions of Grillparzer and Catulle Mendéz demonstrate. This is but one simple illustration of how the Roman dramatists could re-stage old myths and yet constantly invite the audience to something new. The emphasis upon the interpretation rather than upon the plot is precisely the same as it was in the days of Racine and Shakespeare. How far the Roman dramatists were indebted to predecessors for their very striking employment of song is still a moot problem. Leo,[9] following a suggestion of Crusius, held that the Plautine cantica followed the manner of the contemporary music-hall lyrics of Greece as illustrated by the then recently discovered “Grenfell song.” This theory was rejected by Fraenkel[10] because he found no vital similarity between the Grenfell fragment and the Plautine cantica. In his view the Roman predecessors of Plautus—Livius and Naevius—who paraphrased both tragedy and comedy, had probably developed the cantica in tragedy from Euripidean models and then employed them in comedy as well. This theory has a certain plausibility but cannot yet be tested because the cogent examples of cantica in tragedy must be drawn from Ennius, who was not a predecessor but rather a tardy contemporary of Plautus. The view of Leo has received some little support from a brief and peculiar mime-fragment of the British Museum recently published by Milne.[11] However this fragment is so late that it may represent post-Plautine developments, and therefore cannot be pressed into decisive service. It must also be added that recent studies tend to show that Greek New Comedy of the time of Menander had not wholly given up the use of strophic lyrics,[12] and that the Plautine and Ennian cantica themselves seem to have retained not a few traces of strophic structure.[13]