feeding-places and streams, or even from her crofts and streams. In Pindar we find χόρτος λέοντος: in Eur. Iph. T. 134, χόρτων εὐδένδρων. (Hermann's ἀπὸ χλωρῶν πετάλων ἐγρομένα is audacity out of place.)
70. If δειμαίνουσα be right, we have to join εἴτις with γόεδνα . . . in the sense of "I lament whether," which is certainly unnatural. Dindorf prints δεῖμα μένουσα, I suppose merely to show what the metre requires.
78. For the old reading βωμὸς Ἄρης φύγασιν, which is certainly wrong, an obvious correction is βωμὸς ἀρειφύγασιν, which would be unexceptionable if we found it in the text. But other possibilities occur: thus the poet may have written Ἀρησφύγετον in imitation of the word Κρησφύγετον.
80. Εἰ θείη Διὸς εὖ παναληθῶς! This is nonsense. To change Διὸς to θεὸς has no plausibility; all remains abrupt; θεὸς followed by Διὸς is scarcely possible. My present belief is that we get the poet's sense by Ἰθείᾳ Διὸς ἐν παναληθεῖ . . . . which means, "In the straight line of Jove, though drawn with perfect accuracy, the heart's desire of Jove is not easy to trace." Then the abruptness vanishes and the argument is solid. Moreover, in the next line τοι seems to introduce a general maxim. This suggests to me that πάντᾶ, which is weak, ought to be βροντά.
93. τὰν ἄποινον (the old text) is manifestly indefensible. Critics do not seem to have observed that in place of it we need the accusative after ἐξέπραξεν. Then we must put a full-stop after ἐξοπλίζει, and the sense needed will be given by Πάντα νόον δαιμονίων | μνῆμον ἄνω φρόνημά πως | αὐτόθεν ἐξέπραξεν . . .
104. The little word καὶ offends me, and suggests that καὶ διάνοιαν should possibly be παιδὸς ἄνοιαν. Is the spondee in the third foot satisfactory?
107. Remove the stop and join ἐμπρεπῆ with με.
115, 116. πελομένων καλῶς | ἐπιδρομῶς ὅθι θάνατος ἀπῇ. This is mere chaos. The general sense needed is, that
