rn, ro, which have o (churl, learn, earth) and after w (swerve, work, dwarf, sword). The only exception is birch. Hearth, which should be pronounced o°, is occasionally pronounced a. vi. For O.E. long vowels ā, āē (and ie, see iv.), only when these are treated according to the analogy of the corresponding short vowels. Thus ask falls into the analogy of O.E. words with a + s + consonant; last and blast follow fast, mast, and the like; lark follows ark, spark, and others with éa. Wrath owes its a sound to the r (cf. wroth, where the regular sound appears). Ant is an isolated contraction. Darling and farthing owe their sound to the r, but compare forty and thirteen, which are also dissyllables, with eo in O.E. It thus appears that this sound is due directly or indirectly to r in forty cases, to l in five cases, to fricatives in twenty-one cases, and once to a contraction. But not only produced the sound a from vowels; it assumed this vowel sound itself, in some parts of England and the United States, wherever it was final or medial before a consonant, though r and rr before vowels always preserved the consonant sound. This new vowel coalesces with the preceding vowels in such words as mar, star, yard, and is suffixed to the preceding vowel in such words as sheer (shia), hare (heaa), wire (waia), four (foaa), hir (hoa), heard (ho ad). Sometimes the vocalized r forms a syllable, as, for instance, in shower (shaua), ever (eava). When carelessly spoken this vocalized r has the diphthongal sound oa, while in many parts of the United States and of England it has preserved its consonantal character. 4. The tendency of the language in the New English time is not merely a continuation and completion of the process we have already seen in the Ormulum. The tone-lowering had modified in M.E. almost all the classes of words in which it appears in N.E.; but the tone-raising which has affected so large a majority of the O.E. a-sounds in N.E. did not begin till much later, for there is no trace of it in Orm. The influence of the consonants is therefore in some cases earlier and more lasting than in others. Those consonants which have exercised the strongest and most prolonged power have usually affected the entire vowel system in a manner analogous to that in which they modify a. Thus r and h attract all vowels, light and dark, toward a; r + consonant and 7 + consonant usually have a lowering and blunting effect on preceding vowels fricatives lower the tone of bright vowels and blunt the articulation of the darker ones; mutes and nasals agree in their tendency to lower light vowels, to raise dark vowels, and to blunt the articulation of the extreme vowels at either end of the scale. The cause of the modification of the vowel sounds in these cases lies in the relation of the position of the vocal organs when sounding the vowel to that which they must assume to utter the consonant. In speaking rapidly, the former, which may vary, will adapt itself to the latter, which is relatively constant. The liquids and the aspirate h have a near relation to a, which accounts for their influence; less marked, but of the same nature, is the power of the fricatives; while the mutes, being further removed from relationship with the vowels, usually blunt the preceding vowels and assimilate them to the vowel articulation from which the mutes are least removed. In O.E. the general lines of development are similar, except that epenthesis played a most important part, and that the influence of nasals was more distinctly lowering than now. The vowel of the syllable which follows the accent has also some times the power of assimilating the root-vowel, or at least of bringing it nearer to its own sound, which is allied to the form of epenthesis called umlaut. It is difficult to account for these O.E. elements in phonetic growth unless we take them to be a reflection in language of the restless spirit which sees that which is to come as though it were already present, and, while it utters one vowel, "o'erleaps itself and falls on t'other." V. On the Use of the Aorist Participle in Greek. By T. D. SEYMOUR, PROFESSOR OF GREEK IN YALE COLLEGE. UNTIL the present century's historical and comparative study of the Greek language, and even nearly to the middle of this century, it was held that the aorist was the absolute preterit, the expression of a past which is conceived as a unit. The aorist in all its moods, optative, infinitive, and imperative as well as participle and indicative, was thought to refer properly to past time, as truly as does the perfect tense to a completed state or action. In the indicative the aorist was recognized in its full peculiarity. Philological study of the elements of the verb has made it plain that the aorist indicative owes its reference to past time to the augment alone; it has been shown that there is nothing in φυγείν or φυγών, in λιπεῖν oι λιπών, rather than in φεύγειν οι φεύγων, λείπειν or λείπων, which should represent an action as past. We should then expect the aorist participle to be distinguished from the present participle only as the aorist infinitive differs from the present infinitive; i. e. to represent the action as momentary or indefinite, without reference to its continuance. We find in Homer distinct examples of the survival of this usage: e. g. Ω 713 ρ 582 a 23 πᾶν δ ̓ ἦμαρ φερόμην, ἅμα δ' ἠελίῳ καταδύντι ὣς τότε μὲν πρόπαν ἦμαρ ἐς ἠέλιον καταδύντα ἅμα δ' ἠελίῳ καταδύντι πυρσοί τε φλεγέθουσιν ἐπήτριμοι. In these verses the aorist participle seems to be used simply to mark the setting of the sun as a momentary action; the darkness of evening appears to come upon us far more suddenly than the light of morning. This achronic use of the participle is clear also when it is connected with a finite verb in the aorist. Some have thought that in such cases the participle was attracted to the tense of the verb; but this attraction is not easily explained, least of all when the finite verb is not in the indicative; as, Hom,σ 379 τῷ κε ἴδοις πρώτοισιν ἐνὶ προμάχοισι μιγέντα. In such cases it is clear that the action represented by the participle coincides in time with the action represented by the finite verb. A paper read before this Association in 1877 called attention to the temporal coincidence of the aorist par 1 See "Proceedings" for that year, pp. 4, 5. ρ ticiple with the principal verb in sentences like Hom. θ 564 ἀλλὰ τόδ' ὥς ποτε πατρὸς ἐγὼ εἰπόντος ἄκουσα | Ναυσιθόου κτλ. The author of that paper extends his observation not merely to Hom. p 492 τοῦ δ ̓ ὡς οὖν ἤκουσε περίφρων Πηνελόπεια | βλημένου ἐν μεγάρῳ, but also to expressions like Hom. ν 58 καί μιν φωνήσας ἔπεα πτερόεντα προσηύδα, where the principal verb is not in the aorist but in the imperfect. That such cases are frequent is undoubted, but there are many others where the aorist participle refers to an action which precedes that of the principal verb. It is worth an effort to discover the principle which underlies these diversities. We must remember that the old view of the aorist infinitive also was that it properly referred to past time, while now we distinguish its meanings according to its correspondence to an indicative or an optative in the direct discourse. E. g. Plato, Gorgias 449 C καὶ γὰρ αὖ καὶ τοῦτο ἕν ἐστιν ὧν φημι, μηδένα ἂν ἐν βραχυτέροις ἐμοῦ τὰ αὐτὰ εἰπεῖν . . . καὶ οὐδενὸς φήσεις βραχυλογωτέρου ἀκοῦσαι. This is virtually equivalent to οὐδεὶς ἂν εἴποι κτλ. and οὐδενὸς βραχυλογωτέρου ἤκουσας. Thus in the original sentence ἀκοῦσαι naturally refers to past time; the tense is not changed in passing into the oratio obliqua. So e. g. in Isoc. IV 147 ὥσθ' ὁ βασιλεὺς . . . συλλαβεῖν ἐτόλμησεν, we might have had συνέλαβε οι συλλαβεῖν τολμῆσαι. Let us study the participles with the aid of this criterion of correspondence to the indicative or other moods respectively. With regard to the first grand division, the attributive participles, little need be said. It will be admitted readily that as οὗτος (or ἐκεῖνος ὁ λύσων is a rough equivalent of οὗτος (or ἐκεῖνος) ὃς λύσει, and ὁ λελυκώς to ὃς λέλυκε or ὃς λελύκει, so ὁ λύσας corresponds to ὃς ἔλυσε. A Greek would as soon think of confounding λέλυκε and ἔλυσε (and examples of this confusion are rare before the Alexandrian period), as of put 1 Professor Merriam; see his "Phaeacians of Homer," p. 247. 2 Which can be understood when she was told that,' etc.; cf. Hom. N 521 οὐδ ̓ ἄρα πώ τι πέπυστο βριήπυος ἔβριμος ̓́Αρης υἷος ἑοῖο πεσόντος ἐνὶ κρατερῇ ὑσμίνῃ. ting ὁ λελυκώς in place of ὁ λύσας. There is in this aorist participle a reference not merely to past time but to indefinite past time. The aoristic use is clearly developed. But the attributive participle may correspond to the optative or the subjunctive and then the reference to past time is lost. Cf. Hom. Γ 138 τῷ δέ κε νικήσαντι φίλη κεκλήσῃ ἄκοιτις with Γ 71 ὁππότερος δέ κε νικήσῃ κτλ. Ψ 656 τῷ δ ̓ ἄρα νικηθέντι τίθει δέπας αμφικύπελλον. So Plato, Gorgias 458 Α (ἐγώ εἰμι) τῶν ἡδέως μὲν ἂν ἐλεγχθέντων, εἴ τι μὴ ἀληθὲς λέγω, ἡδέως δ ̓ ἂν ἐλεγξάντων, εἴ τίς τι μὴ ἀληθὲς λέγοι, i. e. τούτων οἳ ἐλεγχθεῖεν ἄν ... . . ἐλέγξειαν ἄν. These cases of correspondence to any other mood than the indicative are comparatively infrequent. ... Of the predicate participles, the so-called supplementary participles show most clearly their equivalence to a finite mood. Reference has been made already to passages like Hom. θ 564 ἀλλὰ τόδ' ὥς ποτε πατρὸς ἐγὼν εἰπόντος ἄκουσα | Ναυσιθόου. The direct statement might be: ὁ μὲν πατὴρ εἶπεν ἤκουσα δ ̓ ἐγώ. If in N 521 οὐδ ̓ ἄρα πώ τι πέπυστο . . . υἱος ἑοῖο πεσόντος, πέπυστο had been followed not by a participle but by a declarative clause (its grammatical equivalent), this clause would have taken its verb corresponding to πεσόντος in the aorist, i. ε. ὅτι ἔπεσε υἱός. Compare P 641 ἐπεὶ οὐ μιν ὀνομαι οὐδὲ πεπύσθαι | λυγρῆς ἀγγελίης, ὅτι οἱ φίλος ὤλεθ ̓ ἑταῖρος. The supplementary participle in general retains the full force of its tense after verbs of seeing, knowing, and the like. Isoc. V 62 αἰσθόμενος δ' ̓Αγησίλαον . . . εἰς τὴν ̓Ασίαν διαβεβηκότα καὶ πορθοῦντα κτλ., i. e. ὅτι διαβεβήκει καὶ ἐπόρθει. Pind. Isth. VII 29 ἴστω αὔξων, i. e. ὅτι αὔξει. Soph. Ant. 460 θανουμένη ἐξήδη, I knew that I was to die. So Aesch. Prom. 956 οὐκ ἐκ τῶνδ ̓ ἐγώ ] δισσοὺς τυράννους ἐκπεσόντας ᾐσθόμην ; i. ε. ἐκ τῶνδε περγάμων δισσοὶ τύραννοι ἐξέπεσον. Hom. Ο 422 "Εκτωρ δ ̓ ὡς ἐνόησεν ἀνέψιον ὀφθαλμοῖσιν ] ἐν κονίησι πεσόντα. Bacchylides fr. 6 ξανθότριχα μὲν Φερένικον . . . εἶδε νικάσαντα. Isoc. VI 83 συνειδότες Αθηναίοις ἐκλιποῦσι τὴν πόλιν, i. e. ὅτι οἱ ̓Αθηναῖοι τὴν πόλιν ἐξέλιπον. Isoc. VII 66 τίς οὐ μνημονεύει τὴν δημοκρατίαν οὕτω κοσμή σασαν τὴν πόλιν ; i. e. ὅτι ἡ δημοκρατία οὕτως ἐκόσμησε τὴν |