¹ To mention only a few: Janos Lotz, Das Ungarische Sprachsystem (Stockholm, 1939); Robert A. Hall Jr., An Analytical Grammar of the Hungarian Language (Baltimore, 1938); or, in non-phonemic terminology, J6zsef Szinnyei, Ungarische Sprachlehre. Furthermore cf. Gyula Laziczius, A phonologiar61, Magyar Nyelv 26.24 (1930). The so-called ‘second e phoneme#x2019; is dialectal and irrelevant to this discussion.

This paper has had the benefit of suggestions from George Herzog and Robert A. Hall Jr., to whom my sincere thanks are here expressed.

² This was the substance of my paper read at the 1941 meeting of the Linguistic Society under the title Correlations in the Hungarian Vocalism, which appeared in the Quarterly Journal of Speech 28.449-52 (1942). Since the presentation of that paper I have come across the following statement of Lotz, Ung. Sprachsystem 26: ‘In vielen Dialekten kommen die langen Vokale der hochsten Zungenstellung sehr selten vor . . . . Die einen solchen Dialekt sprechen, haben in der hochsten Reihe nur kurze Vokale.#x2019; In my opinion this is not so much a matter of dialectal but of age variation. I have tried to show that the length phoneme does not occur in the speech of the younger generation of Budapest. But at any rate the variation is free.

³ ‘Probleme der Phonologie, Ungarische Jahrbi.icher 15.495-510 (1936). For Trubetzkoy's discussion of this point cf. Grundzi.ige 26-7 for the important theoretical issue involved.