où (=of-ov) on the analogy of vú. The pronoun où was primarily used for both singular and plural like the Latin sui, its cognate. From opív after the analogy of μé and σé was formed opé also used for both singular and plural. We find this σpé for the dual opwé in Il. 11. III and 115; Od. 8. 271; 21. 192 and 206; Hes. H. S. 62. For both σφώ and σφωέ the possessive is σφωίτερος, which is of you two in Il. I. 216. Just as σφε is used for σφώ or σφωέ, so σφωίτερος is used for opérepos by Apollonius Rhodius in 1. 643 and with reflexive force in 3. 600. It is used for oós (Ap. Rh. 3. 395), just as opérepos is used for the singular as well as the plural. We have what looks at first glance like an interesting parallel to this use of opûï for 'you two' in the Aeneid. We read : (Aen. 3. 493-4.) Vivite felices, quibus est fortuna peracta Iam sua: nos alia ex aliis in fata vocamur. Servius's explanation is satisfactory; sua: id est dura, propria Troianorum, ut (Aen. 6. 62) hac Troiana tenus fuerit fortuna secuta. But it is very probable that Virgil, in thus shaping his verse, had in mind the Homeric use of opú for the second person. XVI THE SCHEMA PINDARICUM AND ALLIED CONSTRUCTIONS WE turn to the Schema Pindaricum or Boeoticum. Following the analogy of: Tà La тpéɣe, thinks Riemann, certain poets, and especially Pindar, use the singular of the verb with the names of things in the masculine or feminine plural. Jelf and Riemann seem right in citing as an example of this construction the well-known couplet of Hipponax : δι' ἡμέραι γυναικός ἐστιν ἥδισται ὅταν γαμῇ τις κακφέρῃ τεθνηκυῖαν (Fr. 29, Β.). Gaisford conjectured elow for or, and Bergk adopted this in his edition, because he found it in one of his manuscripts, as he was almost sure to do. His business was to account for the usual reading σTw, rather than substitute for it a reading which some scribe would inevitably introduce in his copy. Eustathius tells us that Homer invented this schema, and cites to prove his point: καμάτῳ δὲ καὶ ἱδρῷ νωλεμὲς αἰεὶ γούνατά τε κνῆμαί τε πόδες θ ̓ ὑπένερθεν ἑκάστου χειρές τ ̓ ὀφθαλμοί τε παλάσσετο μαρναμένοιν ἀμφ' ἀγαθὸν θεράποντα ποδώκεος Αιακίδαο (11. 17. 385-8). We notice that, while in our first example the subject is a single pair, in this it consists of a large number of pairs; both of these might take the dual in Greek. But we noticed that, when a woman used her two hands to wipe her two cheeks, the plural, not the dual, was the number in use, not merely for the verb, but for the pairs connected with it. And we also noticed that in Latin the dual found expression, not always by the plural, but often by the singular; and that there were traces of this usage in Greek as well. To follow the old rule, the number of παλάσσετο may be determined by the nearest subject ὀφθαλμοί, and its synonym ooσe takes the singular thrice in Homer. This view of the construction is strengthened when we turn to : πνοιῇ δ ̓ Εὐμήλοιο μετάφρενον εὐρές τ ̓ ὤμω and we feel it natural to compare this with: οὔτε τοι ὀξύτατον κεφαλῆς ἐκδέρκεται ὄσσε (ΙΙ. 23. 477), and still more when we turn to : εἰ ἔστιν τούτω διττ τω βίω (Plat. Gorg. 500 D), or to: or to : ἡμῖν γὰρ οὐκ ἔστ ̓ οὔτε κάρυ ̓ ἐκ φορμίδος δούλω διαῤῥιπτοῦντε τοῖς θεωμένοις (Ar. Vesp. 58-9), καὶ Μάγος Αραβος, Αρτάβης τε Βάκτριος, σκληρᾶς μέτοικος γῆς ἐκεῖ κατέφθιτο (Aesch. Pers. 318-19), or even to : Όμηρος μέν νυν καὶ τὰ Κύπρια ἔπεα χαιρέτω (Hdt. 2. 118). Wilpert (Das Schema Pindaricum, Oppeln, 1900) quotes as an example of this schema: ἢ τοι ἐμοὶ χλαῖναι καὶ ῥήγεα σιγαλόεντα where we have plainly a duality of collectives; but it is easier to regard the ήχθετο as determined by ῥήγεα, the nearest subject. Still this is not so plausible in: ἵνα δοκοῦντι δικαίῳ εἶναι γίγνηται ἀπὸ τῆς δόξης ἀρχαί τε καὶ γάμοι (Plat. Rep. 363 Α), or in: ἔστι μέν που καὶ ἐν ταῖς ἄλλαις πόλεσιν ἄρχοντές τε καὶ δῆμος (ib. 463 ^), or in : σκέλη μὲν οὖν χειρές τε ταύτῃ καὶ διὰ ταῦτα προσέφυ πᾶσιν (Plat. Tim. 45 A), or in: ἀφ ̓ ὧν ἐμοὶ ξενίαι καὶ φιλότητες πρὸς πολλοὺς καὶ βασιλέας καὶ πόλεις καὶ ἄλλους ἰδίᾳ ξένους γεγένηται (Andoc. 1. 145). In the last example the two subjects are so closely allied in meaning, that we might assume that the singular verb is determined by what Gildersleeve calls a Unity of Subject. But of this more presently. It seems probable that in the examples cited we have a duality of idea in the subject; and that this follows the same rule as do the duals cited above in taking the verb in the singular. So from: σάρκες καὶ νεῦρα ἐξ αἵματος γίγνεται (Plat. Tim. 82 c) we may assume that in: οὐδ ̓ ἔπι χεῖρες οὐδὲ πόδες (Hdt. 6. 86 γ' 2) ἔπι is for ἔπεστι. This tendency to take the verb in the singular was probably strengthened by the construction with the singular of another class of plurals expressing a unity of idea, a class that seems to have fused with the duals cited above in creating this schema. We have such a plural in : and in : ξανθαὶ δὲ κόμαι κατενήνοθεν ὤμους (Hym. in Cer. 279), ἐκ δέ οἱ ὤμων ἦν ἑκατὸν κεφαλαὶ ὄφιος (Hes. Theo. 824-5), where the snakes' heads are felt as constituting a mane. It is felt, we are told, in : τῆς δ ̓ ἦν τρεῖς κεφαλαί (Hes. Theo. 321) that after ἦν a subject is expected designating the head of the monster, and from this we have a singular verb, but it finds expression in τρεῖς κεφαλαί. With a feeling that this was the true explanation of the schema grammarians everywhere quoted: ἐν δ ̓ ἀγαθοῖσι κεῖται πατρώϊαι κεδναὶ πολίων κυβερνάσιες (Pind. Pyth. 10. 71-2), and taught: The speaker begins with a single subject in mind; and this determines the number of the verb, but afterwards finds expression in a plural form. But we read in Pindar : my Wilpert quotes : λάχναι νιν μέλαν γένειον ἔρεφεν (ΟΙ. 1. 68) (where all editions read ἔρεφον). And, to turn to prose, we have in Plato : πάχναι καὶ χάλαζαι καὶ ἐρυσίβαι ἐκ πλεονεξίας καὶ ἀκοσμίας περὶ ἄλληλα τῶν τοιούτων γίγνεται ἐρωτικῶν (Symp. 188 Β). Even when the verb precedes the subject, as in : or: ἦν δ ̓ ἀμφίπλεκτοι κλίμακες ἦν δὲ μετώπων ὀλόεντα πλήγματα καὶ στόνος ἀμφοῖν (Soph. Trach. 520-2), ἐνὴν δ ̓ ὑφανταὶ γράμμασιν τοιαίδ' ὑφαί (Eur. Ion 1146), or : διήγεται σάρκες (Pind. Fr. 183), where σάρκες is the Latin caro, or in : μέγα τοι δύναται νεβρῶν παμποίκιλοι στολίδες (Eur. Hel. 1358), the interval between the verb and the following subject is so short as to make this view improbable. The plural subject σápkes gives us a singular idea, and Wilpert thinks that σroxides designates a single garment. But the theory deduced from the position of κυβερνάσιες does fit the latest form of this schema, and the form usually found in prose. We read in Herodotus: ἔστι δὲ μεταξὺ τῆς τε παλαιῆς πόλιος, ἢ τότε ἐπολιορκέετο, καὶ τοῦ νηοῦ ἑπτὰ στάδιοι (1. 26). So in Plato : ἔστι γὰρ ἔμοιγε καὶ βωμοί (Euthyd. 302 c) and in : ἐγένετο δὲ μετὰ τοὺς λόγους τούτους ὡσεὶ ἡμέραι ὀκτώ (Luke 9. 28). But in: ἔστι δὲ ἑπτὰ στάδιοι ἐξ ̓Αβύδου ἐς τὴν ἀπαντίον (Hdt. 7. 34) there is no interval between the verb and its subject, and the unity of idea in ἑπτὰ στάδιοι may well determine the number of ἔστι. The feeling of interval is very strong in : προσξυνεβάλετο οὐκ ἐλάχιστον τῆς ὁρμῆς αἱ Πελοποννησίων νῆες ἐς Ἰωνίαν ἐκείνοις βοηθοὶ τολμήσασαι παρακινδυνεῦσαι (Thuc. 3. 36. 2), where Arnold thinks that the subject the speaker has in mind, when he begins, is τὸ τὰς ναῦς τολμῆσαι. In this prose form Kühner seems justified in comparing the schema with the French: il est des hommes, or : il est cent choses, where the German seems more logical with its : es sind viele Dinge. But to account for the origin of the schema from this last stage in its history, as is so commonly done, seems absurd. Even here the original duality of the subject is often evident, as in: v δὲ τοῦ δανείσματος τετταράκοντα μὲν καὶ πέντε μναὶ ἐμαί, τάλαντον δ' Εὐέργου (Dem. 37. 4). Our reference to il est leads us naturally to the use of orw of for ἔνιοι, as we see it in: ἐνταῦθα δὴ οἱ βάλλοντες ταῖς βολαῖς ἔστιν οἳ καὶ ἐτύγχανον καὶ θωράκων καὶ γέρρων, οἱ δὲ καὶ μηροῦ καὶ κνημῖδος (Xen. Cyr. 2. 3. 18). We find ἔστιν ἅ for ἔνια in: καὶ (Κλεόπομπος) ἀποβάσεις ποιησάμενος τῆς τε παραθαλασσίου ἔστιν ἃ (χωρία) ἐδῄωσε καὶ Θρόνιον εἷλεν (Thuc. 2. 26. 2). Ἔστιν @ for ἔνια is quite in harmony with Greek syntax, suitable in meaning, and of a like number of syllables. The natural result of this correspondence would be that in Tw a the primary meaning of the first two syllables would be obscured to some extent, just as in via the primary force of evɩ- was partially lost, and they passed from 'some other' (cf. Skt. anye) to 'some'. In σTw apparently it was the idea of number that was obscured, and this seems to have taken place gradually; and by a series of stages of which σT of was the last. The idiom belongs to prose, and appears first in Herodotus. But neither Herodotus nor Thucydides gives us σTwo, which we find first in Xenophon. Herodotus has : εἰσὶ δὲ οἱ λέγουσι τοὺς ἀπ ̓ Αἰγύπτου νικῆσαι Πολυκράτεα (3. 45); and Thucydides: εἰσὶ δὲ αἳ καὶ οἰκήτορας μeтéẞadov áλioкóμevaι (1. 23. 2) et saepius. But in cases where the masculine and feminine of the pronoun do not differ from the neuter in form, we find the transition to them already in Herodotus and Thucydides; e. g. in: προδοῦναι τὰ ῥέεθρα τῶν ποταμῶν ἔστι ὧν (Herod. 7. 187); and: auxμoi te čσTI Taρ' ois μeɣáλoi (Thuc. 1. 23. 3); and so in Plato : ἔστιν ὅτε καὶ οἷς βέλτιον τεθνάναι ἢ ζῆν (Phaedo 62 A); and in Xenophon: σTw ois kaì πávν åρéσкeι; (Mem. 2. 3. 6). But in Plato we find masculine forms of the pronoun differing from the neuter thus constructed, as in: ἄκων δ ̓ ἔστιν οὓς ἐγὼ ἐπαινῶ καὶ φιλῶ (Prot. 346 E) and in Xenophon: ἔστι δὲ ἂς ἂν καὶ πόλεις τῆς ἀναγραφῆς ὀρεγομένας (Vect. 3. 11). Last of all we come to ἔστιν οἵ in the example first cited; but even in later Greek cioèv of seems more usual. Propertius has imitated or ois in: |