286 



NATURE 



[June 7, 1917 



studied from such a tracing- as this. Such are : 

 (i) the extent of " aspiration " of the plosive con- 

 sonants (shown by the distances between the ver- 



Fk;. s-^Mouth-tracings of l>te, bead, bean, bjaf, bid, bin, bit, showing lengths of vowels and final 



consonants. 



tical Unes i and 2, 4 and 5, 7 and 8); {2) the 

 leng-ths of the vowels (shown by the distances 

 between the vertical lines 2 and 3, 5 and 6, 8 and 

 9). The variations in the 

 pitch of the voice may also 

 be calculated to any degree 

 of accuracy by measuring 

 the voice vibrations in suc- 

 cessive small intervals. 



Fig. 5 illustrates the 

 variations in length which 

 English vowels undergo 

 under certain conditions. 

 The first four tracings show 

 variations in the length of 

 the English sound of ee as 



exhibited in the words hee, head, hean, beat, and 

 the remaining three tracings show similar varia- 

 tions in the length of the so-called " short i " in 

 the words hid, hin, hit. It will be seen that the 



hid is just about the same length as the so-called 

 " long " vowel In heat. (Ignorance of the fact 

 that the vowels in words like heat, late, are much 

 shorter than those in head, laid, 

 is the cause of noticeable mis- 

 pronunciation on the part of 

 many foreigners.) 



Fig, 6 shows (A) a mouth- 

 tracing of the word play said by 

 the writer, (B) a mouth-tracing 

 of the same word said by a 

 Flemish-speaking Belgian with a 

 bad accent. It will be noticed 

 that the Belgian mispronounced 

 the I by making it completely 

 voiced ; in normal English this I 

 is partially devocalised, i.e. the 

 vibration of the vocal chords does 

 not begin until an appreciable 

 time after the explosion of the p. 

 Fig. 7 is a record of good 

 morning (as said on parting), in 

 which tracings of the nose,, 

 mouth, larynx, and a time- 

 measurer have been taken simul- 

 taneously. The points at which 

 the various sounds begin and 

 end are clearly seen from the 

 nose and mouth tracings. (This 

 is where kymographic tracings have an advantage 

 over enlargements of talking-machine records.) 

 The distances between the vertical lines show the 



Fig. 7.— Simultaneous nose, mouth, and larynx tracings oi good morning {a% said on parting). 



vowels in hean and heat differ from that in head 

 in somewhat the same manner as the vowels in 

 hin and hit differ from that in hid. It will also 

 be observed that the so-called " short " vowel in 

 NO. 2484, VOL. 99] 



Fig. 6.— a, mouth-tracing of play as said by the writer ; B, mouth-tracing of the same word 

 mispronounced by a Flemish-speaking Belgian. 



lengths of the various sounds. From the nose- 

 tracings we may gather information as to the 

 extent to which nasal consonants exert a nasal- 

 ising influence on neighbouring vowels. The 

 larynx-tracing shows vibration- 

 waves throughout, since every 

 sound is voiced ; this would be 

 the most convenient curve to 

 use for the purpose of calculat- 

 ing pitch. 



Fig. 8 shows mouth-tracings 

 of the English buckle and the 

 French houcle. Two import- 

 ant differences will be noticed 

 in regard to the consonants : 

 (A) the English I is voiced, 

 whereas the French I is not; (B) in the French 

 word the fe-sound is held on about twice as 

 it is in the English word. The 

 of the voice-waves in the French 



long as 

 smallness 



