228 Proceedings of Boyal Society of Edinburgh. [sess. 
resonance is / 4 , the length of passage is about 50 mm.; an octave 
lower, it is about 100 mm. : the lower octave and a half are ob- 
tained by bulging, as in /. In the upper octave the difference be- 
tween an f passage, a 6 passage and a x passage is very simple and 
obvious : the 0 passage sharply converges ; the x passage sharply 
diverges, and the /passage does neither. After the changes of this 
upper octave have been noted in the x and x passage by the eye, 
those of the 0 passage are readily imagined, for the same octave, by 
conceiving a reversal of the passage, end for end. The cause which 
creates in practice a certain gap between x and x is the organic 
facility of creating a frictional passage, either (1) just behind the 
alveolars, or (2) between the post-dorsum and the velum. The so- 
called “ front ” vowels, such as i and e, connect themselves most 
easily with the former ; hut the “ hack ” vowels, a, o , u , with the 
latter. Thus it is the vowel in licht and lodi which determines the 
consonant, making the one x and the other x- 
A question here suggests itself to the investigator, which must 
he considered before going further. The resonances of / and 0 
came from behind the constriction. How do we know that no 
resonance comes from behind the constriction of x and /! To 
this question it is impossible to answer quite positively. The ear 
does not recognise any such secondary element. But a weak 
secondary element might conceivably he there, and he essential to 
the x and though not separately audible. In whispered fricatives 
there is such a secondary element, which at first escapes separate 
recognition, hut may easily he felt by comparing / with whispered 
v, or 6 with whispered <T, or any similar pair. The secondary re- 
sonance in x or x, if it exists, must he weaker still, for it eludes 
the ear under all circumstances. 
Moreover in x and x , the state of things inside the constriction 
is not what it is in / and 0. Inside the constriction of / and 0 it is 
essential that a resonant tube shall have been framed ; inside that 
of x or x the formation of such a tube seems to be sedulously 
avoided. The observer, feeling inside the x constriction, is sur- 
prised to find how abruptly it terminates inwardly : the tongue 
slopes down as nearly perpendicularly as it can. In the higher 
pitches of x, as the passage outside the constriction grows shorter 
and shorter, this abrupt decline of the tongue inside the constric- 
