234 Proceedings of Boycd Society of Edinburgh. [sess. 
But even when the inner constriction, and what lies behind it, 
is examined, the essential difference between s and f escapes 
observation at first. It is easy to discover that in both of these 
consonants (as in / and 6, but not as in x and f) there is a 
Tesonant passage behind the constriction. This passage is less 
accessible to observation than any such passage that has been yet 
oxamined. But it is evidently subject to exactly the same kind 
of modifications which have been seen in all the other cases. The 
resonance can he made to move, by modifications of the passage, 
through two octaves of pitch. And that the modifications are of 
the same kind as before, is evidenced externally, at the upper end 
of the scale by the same tension in the muscles beneath the jaw, 
and at the lower end of the scale by the same vigorous rounding 
and protrusion of the lips. Internally also the lengthening and 
shortening of passage which creates the upper octave of resonance 
can be directly observed; and though the bulging of passage 
which creates the lower octave is less accessible to direct observa- 
tion, the same muscular feelings are experienced as in former 
cases. Therefore when the resonance is found to extend through 
two octaves, from g to g , we conclude as before that the real range 
is from p 4 3168 v.d. down to g 2 792 v.d. We expect a slight rise 
above previous resonances, because the passage has again been 
foreshortened by a few millimetres, through the retreat of its 
orifice from the teeth to the alveolar ridge, and the result stated is 
consistent with that expectation. 
Up to this point there is nothing to indicate any difference 
between s and f. There are accidental differences to be noted 
presently ; but just as in / and 0, so here no essential difference 
is to be found between those two consonants, except in the nature 
of the frictional exit from the resonant passage. In f this orifice is 
framed on a much larger scale than in s. It is not only wider, 
but it ^extends much further back. The s orifice is as short a one 
as the tongue can possibly frame. It is also framed against the 
ridge of the alveolars ; so that the inner passage widens immedi- 
ately into non-frictional dimensions. But if this s is changed 
into f the tongue at once arches itself upwards, so that the whole 
fore-tongue is brought close to the upper slope of the alveolars, and 
thus a frictional channel far longer than that of s — longer even 
