To be satisfactory, a scholarly interpretation of a Greek tragedy must enable the present-day reader to see the play, so far as is possible, through the eyes of the fifth-century audience. If it does not, if it merely substitutes the predilections of a particular scholar for those of the reader, it is useless, and indeed worse than useless; for the reader unassisted by the interpretation of others may well examine the play critically for himself, while the reader with an interpretation at his elbow is likely to make every effort to fit the ideas of the tragedian into the schema provided for him. Certainly, more and more interpretations are of the kind which assist the reader; but a significant number even of the most recent throw more darkness than light on their subject by refusing to acknowledge the whole context of value and belief in terms of which the tragedians wrote, and the audience watched, these dramas. The result of such misinterpretation is frequently, as in the case of the two I shall discuss here, to present for our admiration a more high-minded and uplifting drama than the one which (say) Euripides wrote; but our concern is presumably not to achieve this end, but to understand Euripides and his audience.

In Part I of this article the major problems of the transmission of the Bacchae were considered, with a discussion of interpolated lines and lacunae, whether certain or merely postulated by previous editors. In the Introduction it was argued that P is a copy of a manuscript which was very like L (whether a copy or a twin hardly matters) before being supplemented with variant readings and with the whole of Tr. and Ba. 756 ff. from a lost source. The symbols λ and were used for P's exemplar and this remoter source respectively. The ancestry of in turn was traced back to Λ, the ancestor of L. This, if right, means that L and P (including the end of Ba.) both descend from a single uncial archetype, though with a fair chance that some true variants from a different tradition reached and λ. P, however, steadfastly preferred the original reading of λ (= L, in effect) whenever possible. When P differs from L, there are three possibilities: (a) that he is making a fresh error (misreading or carelessness—seldom, if ever, constructive); (b) that λ (before being supplemented) differed from L (probably only in ortkographica); (c) that λ (after contamination) included a variant which P preferred (significant variants are so rare that legibility is likely to have been P's main criterion).

In the last number of CQ,(N.S. xvi [1966], 28 [iii]), I mistakenly attributed to Professor Zuntz (An Inquiry into the Transmission of the Plays of Euripides [Cambridge, 1965]) the view that P's exemplar was, before correction, the ‘parent’ of L.

The possibility that the parent (‘Λ’ in Zuntz) of L became, after corrections derived from its own parent (‘β’), the parent of P is in fact considered by Zuntz (p. 124) only as an alternative to the hypothesis, propounded by him on p. 122 and set out in his stemma on p. 192, that the parent (‘π’) of P was a copy of Λ and thus a brother of L. Regrettably, An Inquiry appeared too late for me to do it justice, with my article already in proof. In addition, I now draw attention to the following passages in An Inquiry as containing arguments which affect what I have written: pp. 13 ff. (my p. 28, n. 1), 119 f. (on Ba. 144–7), and 118 (on Ba. 151).