¹ “In French, the breath-stream is interrupted after a consonant and before a following consonant, after a consonant and before a following vowel: …” (“Beginning-Consonants and Breath-Control in French and English,” PMLA, xlix 1181.) “Before each consonant and between the explosion of the consonant and the beginning of the following vowel, there are ‘breaks’ in the air current, and the consonant represents the beginning or the zero point in a new movement.” (“Rate, Direction, and Continuity of Movement in French and English Speech,” PMLA, xlv, 1261.) “Very different is the French pronunciation of cette tante. … The expulsion of the air is stopped after pronouncing [e] and before the tongue takes the position of closure for [t], then the effort to expel the air is resumed again after the tongue has made contact with the teeth and gums for [t], and, to permit of the explosion after the [t], necessary to pronounce it, the tongue is lowered. After the explosion, the expulsion of the air is stopped again …, the tongue makes contact again with the teeth and gums, and the second [t] is produced as the expulsion of the air is resumed.” (“An Explanation of the Differences in Length and Voicing of Consonants in French and English,” M P, xxvi, 342.)

² The fact that many of Mr. Barker's kymograms do not show voice vibrations during the transitions loses its significance when it is observed that the tracings frequently fail to show vibrations for voiced consonants and even for vowels.

³ C. E. Parmenter and S. N. Treviño, “A Technique for the Analysis of Pitch in Connected Discourse,” Archives Néerlandaises de Phonétique Expérimentale, vii, 12.

⁴ “Beginning-Consonants and Breath-Control in French and English,” PMLA, xlix, 1168.

⁵ “End Consonants and Breath-Control in French and English,” MP, xiv, 421–422. On tombe see “Beginning-Consonants and Breath-Control in French and English,” PMLA, xlix, 1167.

⁶ “Beginning-Consonants and Breath-Control in French and English,” PMLA, xlix, 1181.

⁷ “An Explanation of the Differences in Length and Voicing of Consonants in French and English,” MP, xxvi, 342.

⁸ The vertical lines drawn between the tracings in each kymogram are not intended to delimit the sounds. These lines indicate roughly reference points between the spirograph and larynx tracings.

⁹ Idem., p. 342.

¹⁰ Idem, p. 347.

¹¹ “Beginning-Consonants and Breath-Control in French and English,” PMLA, xlix, 1169.

¹² Idem, p. 1167.

¹³ Idem, p. 1177.

¹⁴ Oscillograms of the connected speech of French subjects support the conclusion reached. Because of their length, these oscillograms are not published. One of them was exhibited at the meeting of the Modern Language Association at Philadelphia.

¹⁵ Mr. Barker says that there is evidence of interruption of the breath stream during transitions not only in French but in all the other Romance languages and in the Slavonic languages (MP, xiv, 422; xxvi, 351; PMLA, xlix, 1168). As in the case of French, there is ample experimental evidence for Italian to refute his contention. In L'analisi elettroacustica del linguaggio, 2 vols. (Milan, 1934), Gemelli and Pastori have made available a great number of oscillograms of Italian words and phrases. An examination of the oscillograms reveals no interruption of vibrations between voiced sounds within a group, and the authors say that within a phrase each sound fuses into the preceding and following one (pp. 125, 234.)