that παλινάγρετος is from πάλιν ἀγείρειν, since although no certain instance of ȧyperós = 'collected' is quoted, it is clear that a number of nomina agentis in -aypeτns belong to ἀγείρω (e.g. ἱππαγρέτης, μαζαγρέτης and others collected by E. Fraenkel, Glotta I. 281; Nomina Agentis I. p. 56), and the nomina agentis in -rs often correspond in form to the verbals in -TOS. But the meaning of παλινάγρετος in Homer (οὐ γὰρ ἐμὸν παλινάγρετον . . . ὅτι κεν κεφάλη κατανεύσω) undoubtedly points towards ἀγρέω.

In -ayperós, then, we have a word whose meaning connects it with dypéw, while its form connects it also with dyeípw. Remembering similar cases, such as that of ὤφλον (ὀφλεῖν) which belongs in point of sense to ὀφλισκάνω, ὀφλήσω, and in point of form also to opeíλw, one may perhaps suggest that the paradigms of dypéw and ¿yeípʊ originated in the splitting up of an older paradigm, and that dyperós chosen, taken,' is the point of contact between them, as opλeiv is the point of contact between ὀφλισκάνω and ὀφείλω.

The meanings collect,'' select,' and 'take' are found elsewhere united in a single word: Grk. ὀστέα λέγειν, ξύλα πολλὰ λέγεσθε, λεκτός chosen, picked; Lat. legere ossa, legere judices; Russian brat' to take,' vy-brat' to select,' so-brat' to collect.'

It must be admitted that the meaning of dyeipw is already pretty different in Homer from that of ἀγρέω. Nevertheless in τ 197: δημόθεν ἄλφιτα δῶκα καὶ αἴθοπα οἶνον ἀγείρας it is virtually an iterative of λαμβάνειν. The function of the κωλακρέται (: ἀγείρω, Boisacq, s.v.) is mentioned in G.D.I. 5497 : λάψεται . . . κωλῆν μίαν ἀπὸ πάντων. Αγύρτης, which is formed from ἀγείρω in this sense (cf. Fraenkel, Glotta I. 281 n. 1; Nom. Agentis I. p. 56 n. 2) may originally have meant 'receiver, taker,' and subsequently have acquired the sense 'beggar' in the same way as dÉKTYS (: δέχομαι) and perhaps αἴτης (· πτωχός Hes., cf. ἐπαίτης· ὁ πτωχός Suidas, προσαίτης v. Solmsen, Laut- u. Verslehre, p. 97 n. 1) if it belongs to ai-vv-μai, ë§-ai-тos.1 Does ἀγείρω mean choose in Hom. A 142: ἐν δ ̓ ἐρέτας ἐπιτηδὲς ἀγείρομεν (cf. A 309: ἐν δ' ἐρέτας ἔκρινεν ἐείκοσι and the phrase ἐξαίτους ἐρέτας ' picked oarsmen ) ?

Αγρέω may have been created from ἀγρ-ετός (redivided ἀγρετός) according to some such proportion as έλα-τός : ἐλάω (cf. ἔλων Hom.) = ἀγρετός : χ. The unthematic inflexion may have arisen thus: éλa-Tós éλá-vτw (Coan imperative, Buck, Dial. § 162) = ἀγρετός : *, whence x = a form like the Lesbian third plural imperative κατάγρεντον.

Αγέτρια.

The following glosses occur in Hesychius: (r) ἀγέτρια· μαῖα, Ταραντῖνοι; (2) ἀναγέτρια· ἡ ταῖς τικτούσαις ὑπηρετοῦσα γυνὴ παρὰ Ταραντίνοις οὕτω λεγομένη, ἣν Αττικοὶ μαίαν καλοῦσι; (3) ὑφαιρέτρια· μαία. W. Schulze, Quaest. Ep. p. 265 n. 5 (who is followed by Fraenkel, Nomina Agentis I. p. 62) derives άyérpia and dvayérpia from ayev, but has not succeeded in making this really convincing. The third of these glosses provides the clue, and suggests that dyérpiα= ȧypéтpia, with loss of the first p by dissimilation from the second. This dypérpia is the feminine nomen agentis to ἀγρέω = ὑφαιρεῖσθαι. It follows that ἀγρέω existed in Tarentine Greek, and therefore probably in Laconian.

Another word which Hesychius interprets by ipaipeiobai (among other interpretations) is dyypiŝew. It should be brought into relation with the Thessalian ἀγγρίζειν. éparypévlev and poάvypeσris. To the ȧvypéw (cf. Buck, Glossary, and § 58c), which these forms imply, ἀγγρίζω is related as αἰτίζω to αἰτέω, μοχθίζω το μοχθέω, etc.

The accent at any rate, if correct, is an objection to Solmsen's view that poσaiTNTŃS

OXFORD.

became poσairns by haplology.

RODERICK MCKENZIE.

(To be continued.)

SUMMARIES OF PERIODICALS.

LITERATURE AND GENERAL.

[These summaries will in future deal only with original work appearing in the periodicals. Reviews of important books will be mentioned in the lists of new books published in the Classical Review.-EDD. C.Q.]

American Journal of Philology. XLI. 3. 1920.

E G. Sihler, Quintilian of Calagurris (An Essay). Deals, inter alia, with Quintilian's leaning to Stoicism, his endeavours to reinstate Cicero's oratory as a model by which to reform the debased speaking of his own day, and his attitude towards the theories of the Greek professional teachers of rhetoric. Francis A. Wood, Names of Stinging, Gnawing and Rending Animals. A list of names, chiefly of insects with connected words. M. B. Ogle, The Lover's Blindness. Traces in classical literature the forms in which the credulous or the disillusioned lover deals seriatim with the physical characteristics of his mistress. Campbell Bonner, The Trial of St. Eugenia. Illustrates by classical examples, and specially one from the Fabulae of Hyginus c. 274, the revelation of her sex by a woman previously disguising herself as a man. Tenney Frank, Tulliana. Six notes on points arising out of passages in Cicero's Letters or connected therewith.

Berliner philologische Wochenschrift. 1920.

...

Aug. 28. K. Koch, Zu Vergil, Aeneis VI. 327: transportare, cross,' cf. Caesar, B.C. III. 25, 2.- Sept. 4. A. Schöne, Immer noch einmal Sall. Jug. 38, 10: read quia atrocissima metuebantur.-Sept. 11. H. Paasch, Zu Horaz O. III. 14 u. 26, offers in 14, 11 Iani iram expertae; cf. Servius on Aen. I. 294, VII. 610. In Ode 26 Chloe is jocularly likened to a besieged city; 1. 7, read arcam and cf. Cic. Ad Att. I. 16, Tibullus I. 5, 60, 67-8.—Oct. 16. L. Mader, Zu Musaios' Hero u. Leandros. In 246, ψυχρὸν for ἔστιν: Dilthey's οἰχόμενον in 5 requires γάμον ἔννυχον in 4: 180-1, read ἐθελήσῃς Toλúavopos.-Oct. 23. R. Philippson, Die Kúpiaι Sóĝai, a detailed treatment. Though some are taken from Epicurus, he is not the collector.-Oct. 30 and Nov. 6. P. Michael, Zu Aristoteles, discusses at length Pol. 1342 a 11 sqq. (1) kovþíčeσdai μeð' ¿dovîs is not pathological. The spectator, influenced by tragic music, advances (see De Caelo, 311 b 14 sqq., a 1 sqq.) from potentiality as aŋTIKÓS to évreλéxela; he reaches Oewpía and its concomitant pleasure (Eth. Nic. 1177 b 19 sqq., etc.). The phrase coincides in essence with Poetic's dovỳ (peyiσrn 4vxaywyia οι εὐφροσύνη) ἀπὸ τραγῳδίας οἰκεία; but musical katharsis is psycho-physical, while poetic is spiritual. (Plato held the pathological view: Rep. X. 606 a; Soph. 226 d-7 e.) Medicinally, ká@aporis involves maintaining as well as restoring health (Metaph. 1003 a 34 sqq., etc). (2) xapàv áßλaß is no answer to Plato, Rep. X. 605 c sqq. On the whole, A. accepts P.'s classes of music (Laws 667 d e, 668 a b, etc.; Pol. 1339 b 42 sqq., etc.), but adds ȧváravois (Pol. 1339 a 16 sqq., etc.) to maidiά as the aim of the Kalaρтikà μéλŋ, which differ from enthusiastic and tragic music.

D

Classical Philology. XV. 3. 1920.

E. S. McCartney, Forerunners of the Romance Adverbial Suffix. A collection of nouns in the ablative with adjectives agreeing, in which the writer thinks the proper force of the noun had so weakened that, like mente (later a mere adverbial suffix), they could be used adverbially. Such are names of parts of the body, etc., pede, animo; words expressing manner, modo, exemplo, etc. Tenney Frank, Vergil's Apprenticeship, III. On V.'s theories of style and his literary development, with a discussion of the 'molle atque facetum' ascribed to him by Horace, S. I. 10. 44. H. W. Prescott, Inorganic Rôles in Roman Comedy. On the use in plays of characters not vitally concerned in the action, such as parasites, cooks, pueri delicati, as in developing or clearing up a situation, providing entertainment, filling up intervals of time required for stage purposes. These generally come from the Greek originals, though they may be due to 'contaminatio' or Roman remodelling. R. B. Steele, Arrian's Anabasis and Book XVII. of Diodorus. Comparison of the writers in regard to 'form of statement,' differences being considered under the heads of personality and diction, including syntax. The data show 'two types of literary expression distinct in many particulars.' The resemblances need no unity of source for their explanation.' In Notes and Discussions E. T. Merrill replies to Tenney Frank's argument (C.Ph. XV. 117) that the 'Sabinus mulio' of Catalepton X. was Ventidius Bassus.

Classical Weekly (New York). 1920.

Oct. 4. H. L. Warren, The Foundations of Classic Architecture (A. D. F. Hamlin). 'The most serious and discriminating contribution made in recent years towards an understanding of the origin, development, and character of classic architecture.' It deals with Egypt, Mesopotamia, Persia, the Aegean, and, more fully, with Greece. 'I know of no clearer or more suggestive presentation of Greek architecture, either as a whole or in detail.'

Oct. II. E. Riess contributes a note on Aeneid vi. 42-44. Norden, in his commentary, quotes a statement in Beloch's Campanien, 161, to the effect that at Cumae the whole rock is perforated by grottoes, arranged in three levels. In an article in The National Geographic Magazine of May, 1920, W. A. Griffiths gives an account of 'a system of subterranean places of worship' discovered in Malta in 1902. He says there are three series of chambers excavated out of the solid rock, on three levels ... we come to a square entrance into a small round cave a yard or two in diameter. Possibly the oracle was kept there. A little farther in the cave, at about the level of a man's mouth, is a hemispherical hole in the side wall about two feet in diameter. It was noticed that any word spoken into this place was magnified a hundredfold and audible throughout the entire underground structure.'

Oct. 18. G. Ferrero and C. Barbagallo, A Short History of Rome, translated by G. Chrystal (G. A. Harrer). The work seems to have too much originality in its treatment of the origins of Rome, to be rather inadequate in the history of the first century A.D., and to present too many details in portions of the later history. . . . The work of translation has been satisfactorily done.'

Oct. 25. E. Nitchie, Vergil and the English Poets (M. B. Ogle). The purpose of this study is 'to trace the changes in the reaction to Vergil's poetry in the different periods of English literature.' It is a large undertaking, . . . but the author has succeeded in gathering the most important facts.'

Nov. 8. A. E. R. Boak, The Master of the Offices in the Later Roman and Byzantine Empires (W. A. Oldfather). This is the "special study" promised by the author in Harvard Studies in Classical Philology 26 (1915), in explaining the omission of the magister officiorum from his elaborate dissertation upon the Roman Magistri in the civil and military service of the Empire. Like the former monograph this has been well planned and thoroughly carried out.'

Hermathena. XLII. 1920.

C. Exon, The Evolution of the Subjunctive Form. Adopting and defending the view of Thurneysen that there was a time when one set of verb-forms did duty for all tenses and moods of the finite verb, E. by reconstructing these primitive forms as they existed in the parent speech seeks to trace step by step the development of the earliest form of the 'subjunctive mood.' W. J. M. Starkie (the late) illustrates the Aristotelian analysis of the comic (which may descend from a lost portion of Aristotle's Poetics) from Aristophanes, Rabelais, Shakespeare, and Molière. J. P. Postgate, On Some Quantities in Phaedrus: (1) Nihil and Nil. To the twelve passages where the word occurs before a vowel in this author P. applies the Ovidian rule 'that nil may be used in the "rise" of a foot before vowels as well as consonants, but in the "fall" before consonants only.' This leads to a preference for nil in nine of the passages and the retention of nihil in the remaining three. Before a consonant the MSS. give nihil in two passages where it is metrically possible. Hence P. will not accept the dictum of L. Muller that Phaedrus does not use nihil except in the final iambus. (2) Vespertilio. P. shows that there is no classical authority for vespertilio. J. G. Smyly, Heron's Formula for Cube Root. Heron's Metrica, recently discovered at Constantinople, provide a single incidental example of the extraction of a cube root. M. Curtze has proposed a formula which, though giving Heron's result in this particular case, is, as he himself admits, very inexact for other numbers. S. shows that it is possible to obtain from Heron's words a different formula-one of astonishing accuracy-which for high numbers gives even better results than can be obtained from a table of seven-figure logarithms.' E. H. Alton has notes on the Culex, also an article Anna Perenna and Mamurius Veturius.' Ovid describes the festival of Anna Perenna (Fast. 3. 523 sqq.) ending with the lines: occurrit nuperuisa est mihi digna relatu-pompa: senem potum pota trahebat anus.' A. suggests that the 'pota anus' is Anna Perenna herself, at least in effigy, comparing Verrius ap. Paul. 243 Petreia uocabatur quae pompam praecedens in coloniis aut municipiis imitabatur anum ebriam.' Petreia may be an Italian sister of Anna Perenna. Usener has shown that Anna had a male companion who he maintains was Mamurius Veturius. The name of the male character may however have varied, but the type was fixed; it represented the useless and foolish old man. J. G. Smyly, Some Examples of Greek Arithmetic. Papyrus 186 (Papiri della Soc. Italiana) contains two arithmetical problems not adequately dealt with by the editors. S. while admitting that it may be impossible to restore the exact words, has reproduced the complete mathematical calculation. R. M. Gwynn, Notes on the Vocabulary of Ecclesiastes in Greek. G. illustrates the book from contemporary Greek 'on lines somewhat similar to those adopted by Drs. Moulton and Milligan for the N. T.' M. Esposito, Classical MSS. in Irish Libraries. I. deals with those in Trinity College, Dublin. W. J. M. Starkie, Matt. xxvi. 45 and xxviii. 2. (1) кαlevdete λοιπὸν καὶ αναπαύεσθε. S. argues that λοιπόν here as in Mod. Gk. = Anc. Gk. δ' οὖν 'will then,' a use not unknown in Ancient Greek. (2) σepòs éyéveтo péyas. S. here takes oeuμós to mean 'a quaking.' No earthquake is mentioned in Mark or Luke.

Hermes. LV. 4. 1920.

A. Rosenberg, Die Entstehung des sogenannten Foedus Cassianum und des latinischen Rechts. The treaty which Sp. Cassius is supposed to have made with the Latins in 493 was probably made about 280, and was part of the efforts made by Roman diplomacy to present a united front in Italy against Pyrrhus. The terms as given in Dionysius are held to show the influence of Greek political thought. R. Philippson, Zu Philodems Schrift über die Frommigkeit IV. P.'s criticism of the earlier philosophers (to be continued). C. Robert, Die Hera von Tiryns. A refutation of Frickenhaus'

view that Tiryns was the earliest seat of the worship of Hera. E. Preuner, Honestos. An elaborate examination of the epigrams of Honestos. The article includes incidentally a discussion of the genealogy of the Attalids. F. Münzer, Die Fanniusfrage. An attempt to explain the meaning of Cicero's question about the Fannii in his letter to Atticus XII. v. 3.

Mnemosyne. XLVII. 4. 1919.

W. Vollgraff, De Theocriti et Callimachi Dialecto, points out that the Doric of Theocritus contains a considerable number of Aeolic forms, and supplies a list of those found in XIV. and XV. The same holds of Callimachus, although not to the same extent. An inscription of Cyrene dating from A.D. 20 similarly reveals a number of Aeolicisms. V. concludes from an examination of this evidence that Callimachus was writing the dialect of his native city, and Theocritus Syracusan. J. van Wageningen, Manere = Esse, collects a large number of instances from Manilius in which this equation seems to hold; also instances where maneve has its ordinary meaning. W. Vollgraff, Theocritea, gives annotations critical and explanatory on Theocr. XIV. and XV. H. Wagenvoort, Obiter Tacta, (1) emends Pliny, Ep. I. 4, to read: 'Exceptis (MSS. ex) epistulis meis (nam iam tuis opus non est: una illa brevis et vetus sufficit) non mehercule tam mea sunt, quae mea sunt, quam quae tua,... the 'brevis et vetus' referring to that in which his correspondent accepted him as suitor for her daughter's hand. (2) Comparing Ep. IX. 10, where P. is generally admitted to be referring to Tac. Dialogus, with Ep. VII. 20, where he talks of having revised some work of Tacitus, W. concludes that the latter passage refers to the Dialogus also, which he infers was finished in or about 108 A.D. (3) Some notes on Seneca Phaedra, (4) on Ovid Ex Ponto IV. 16. 33, which he emends to Tityrus antiquas, Passenne, redisset ad herbas.' Passennus is an Umbrian gentile name, otherwise Passienus. W. thinks Ovid is referring to L. Passienus Rufus mentioned in Vell. Pat. II. 116. A. G. Roos, De Rescripto Imp. Severi et Caracallae Nuper Reperto, offers suggestions for the restoration and interpretation of an inscription recently found on the site of the Roman colony of Aelia Solva near modern Leibnitz; it is dated to the year A.D. 205, and contains the names of members of the guild of firemen and a warning against the richer classes attempting to evade their obligations by becoming enrolled in the guild. P. J. Enk, De Lydia et Diris carminibus, seeks to prove the author of these poems to have been Varius. C. Brakman has notes on Velleius Paterculus. J. J. Hartmann proposes 'Claudius' for 'clausus ab' in Propertius III. 18. 1. J. C. Naber, Observatiunculae De Iure Romano, concludes his. section De Mensurae Generibus, adding a geographical and topographical index to the whole chapter.

Rassegna Italiana. I. 3.

G. Pasquali continues his analysis of the 'Characters' of Theophrastus. Of his followers, (1) Lycon clearly aimed at embodying a' character' in an ethical treatise, and is less genial and convincing than Theophrastus because more rhetorical and long-winded. (2) Satyrus is quoted by Athenaeus in a fragmentary sketch of the äowTo or spendthrifts.' (3) Ariston of Ceos, who to a letter on methods of curing pride adds a series of kindred 'characters,' abounds in subtle distinctions and definitions such as were probably added by Theophrastus verbally in discussions with his pupils. P. denies that Ariston merely aped Theophrastus, and praises his wider outlook. He sets out to prove that the poems and epilogues in the work of Theophrastus are Byzantine additions, finding his proof in the use of the form eidñoa, and in a misunderstanding, natural to a late writer, of several words used by Aristotle and his contemporaries. A. Amatucci quotes a passage from the life of