^{page 11 note 1} 2nd ed., Philadelphia, 1940, pp. 94–105.

^{page 11 note 2} Harmonica Stoicheia, I. 9Google Scholar (101. 19–102. 7, 102. 20–25 Macran); quoted by Sturtevant, op. cit. 96–97. The statements of Dionysios of Halikarnassos, Comp. Verb. 40. 17–41. 12, might be used in favour of the first alternative; but Aristoxenos seems to have a much clearer understanding of accentuation than Dionysios.

^{page 11 note 3} Reprinted by Sturtevant, , op. cit. 101Google Scholar; from Arkadios 186–91 in Lentz, Herodian xxxviii–xxxx; and from Laum, B., Dos alexandrinische Akzentuations-system (Paderborn, 1928), 99.Google Scholar

^{page 12 note 1} These terms are suggested solely for the purpose of this paper.

^{page 12 note 2} This Byzantine use of the grave accent may be merely an orthographic device to indicate the ends of words.

^{page 13 note 1} L'Accentuation des langues indo-européennes (Kraków, 1952).Google Scholar

^{page 14 note 1} Kleine Schriften (Zürich, 1959), 389–98; from Corolla Linguistica, Fest schrift Ferdinand Sommer (Wiesbaden, 1955).Google Scholar

^{page 14 note 2} Slavonic and East European Review, xx (1941), 251–65.Google Scholar

^{page 14 note 3} Ibid. xxxv (1956–7), 409–27.

^{page 14 note 4} For criticisms see Hamp, E. P., Indogermanische Forschungen, lxiv (1958), 40.Google Scholar

^{page 15 note 1} For an account of the rise of Common Lithuanian see Senn, A., Slavonic and East European Review, xxii (1944), 102–16CrossRefGoogle Scholar; and for a description see W. K. Matthews, ibid. xxxv (1956–7), 39–73.

^{page 15 note 2} This statement of the essentials is commonly given by grammarians; it is based on studies made by F. Kurschat. It overlooks some niceties; for example, some long vowels are longer than others, and account of this was taken in the accent-notation proposed by A. Baranauskas. But the usual statement is adequate for most purposes.

^{page 15 note 3} I give in brackets Lithuanian terms used for the three intonations by Salys, A., Lietuviu kalbos tarmės (Tübingen, 1946Google Scholar; I thank my friend Mr. A. Šukys for allowing me access to a copy of this work); for a description of the intona tions compare Pr. Skardžius, , Lietuviu kalbos vadovas (Bielefeld, 1950), 128–9.Google Scholar The English terms, based on those of Salys, are my own, as I am not aware of any recognized terms; for clarity I have avoided using the same terms as for Greek.

Lithuanian grammarians often say that even each syllable without ‘stroke’ has its own intonation, though this is not so easily perceptible as that of syllables with ‘stroke’; it has not been perceptible to me at all.

^{page 16 note 1} Marks of accentuation are only used in grammars, lexica, and similar works.

^{page 17 note 1} Op. cit. 52, 21.

^{page 17 note 2} Ed. Schwyzer, , Griechische Grammatik, i. 382Google Scholar, tries to establish a correspondence between the ‘falling moment’ of Greek and the ‘extended intonation’ of Lithuanian, but the evidence is inadequate.

^{page 19 note 1} The rule was discovered independently by various scholars including de Saussure, F., Indogermanische Forschungen, Anzeiger, vi (1896), 157–66.Google Scholar

^{page 19 note 2} On -ỹbė among the Zanavykai see Salys, , op. cit. 53.Google Scholar On the change within Common Lithuanian see Senn, A., Language, xv (1939), 189–93.CrossRefGoogle Scholar

^{page 19 note 3} Compare akmuõ, liemuõ, stomuõ.

^{page 19 note 4} Plato, Crat. 399a.

^{page 19 note 5} The distinction had already been drawn in the anonymous Universitas linguarum Litvaniae (Vilnius, 1737), but this work was unknown to Schleicher and Kurschat; see Rozwadowski, J. v., Indogermanische Forschungen, vii (1897), 233–70.Google Scholar

^{page 20 note 1} Rhet. 1403^b24–29; Poet. 1456^b30–34. Aristotle alludes in passing to the distinction; so it had been made previously, whether by a predecessor or possibly by himself. His ‘middle’ is not the ‘falling moment’, in view of Soph. Elench. 177^b35–178²3.

^{page 20 note 2} Fr. 84 in the edition of Varro by G. Goetz and F. Schoell (Leipzig, 1910). It is not clear how far the words are precisely those of Varro and how far they are Ps.-Sergius' summary of Varro's views. But Ps.-Sergius seems to have been fully convinced by Varro's exposition; so the whole passage may be attributed to Varro.

^{page 21 note 1} Varro identifies them with βαρεῖα and ⋯ξεῖα respectively; but if Glaukos had meant ‘low’ and ‘high’, he should have said βαρεῖα and ⋯ξεῖα.

^{page 21 note 2} The terms staiginė and smarkinė are used by Skardžius, (above, p. 15, n. 3).

^{page 21 note 3} This is my own impression, which may be unreliable; note, however, that Kurschat, F., Grammatik der littauitschen Sprache (Halle, 1876), 57–68Google Scholar, using musical notation to illustrate the difference between the ‘intonations’, implied a lesser interval for the ‘extended’ than for the ‘pushed intonation’.

^{page 21 note 4} The information that follows is taken from Salys, op. cit. He describes Zhemaitish accentuation on pp. 20–21 and accentuation in northern Aukshtaitish on pp. 59–61. On the ‘broken intonation’ in Latvian see J. Endzelin, Lettisches Lesebuch (Heidelberg, 1922), 3.Google Scholar

^{page 22 note 1} Zhemaitish alters them where they already existed in Protolithuanian; but it develops them subsequently from -ė- and -o-, as in the folk-song quoted below.

^{page 23 note 1} This lament is printed by Senn, A., Handbuch der litauischen Sprache (Heidelberg, 1957), ii. 47 f.Google Scholar Two more examples are given by Leskien, A., Litauisches Lesebuch (Heidelberg, 1919), 33 f.Google Scholar

^{page 24 note 1} Balys, J., Lietuvtu dainos Amerikoje (Boston, 1958)Google Scholar, No. 279. The dialect is Zhemaitish.

^{page 24 note 2} Tremtinys (first published in Würzburg, 1947). This text is taken from A. Senn, op. cit. 87. The metre is: