[172] The compound tone in Swedish, according to Mr. Sweet, “only occurs in words of more than one syllable,” and “consists of a falling tone on the first (the accented syllable), followed by a high tone on the next. The high tone seems to be reached by a jump rather than by a glide. The compound is, therefore, a compound rise distributed over two syllables.” The other Swedish accent, the simple tone, is the negative of the compound one, and answers to the “glottal catch” or stöd tone of Danish.
[173] See Wackernagel in Kuhn’s “Zeitschrift,” 23 (1877).
[174] See Sayce in “Journal of Royal Asiatic Society,” x. 2 (1878), pp. 251, 252.
[175] Sweet: “Handbook,” p. 35.
[176] “Kerenzer Mundart,” pp. 142 sq.
[177] Sweet: “Handbook,” pp. 59, 60.
[178] In many instances, however, guṇa and vṛiddhi seem to be due to the presence of the vowel a in the following syllable, which has been anticipated, as in the case of the German umlaut or the Greek epenthesis (as in λόγοις for the locative λόγο-σι, and then, by false analogy, λόγοισι).
[179] The existence of these velar gutturals was first pointed out by Ascoli, and since by Fick and Havet.
[180] Diez: “Grammatik d. romanisch. Spr.” (2nd edition), i. 248, 254.
[181] “Comparative Grammar of the Anglo-Saxon Language” (1870), p. 28.