236 Proceedings of Eoyal Society of Edinburgh. [sess... 
nants tend more strongly to a fixed type of articulation than any 
previously described ? 
It is simply that there are now two resonances instead of one,,, 
and that these two resonances have to be adjusted so as to rein- 
force each other. It is no longer permissible to arrange the avail- 
able air-space into any shape which will produce a resonance ; it 
must be arranged into two portions whose resonances shall bo 
nearly identical. This is a problem which, were it not for the 
above-named devices of spreading and pursing the lips, and for the 
considerable range which the resonance of the fore-cavity possesses,., 
could only be satisfied by one specific type of articulation. Such 
an articulation would be as fixed as that of a vowel, and for the 
same reason — that the resonances are not severally free, but must 
be more or less accurately adjusted to each other. In English, in 
fact, this seems to be almost the case with s and f ; for all English 
speakers have a strong aversion to employ either rounding or pro- 
trusion of lips in ordinary speech; and, though the range of the 
resonance of the fore-cavity is considerable, there is always a 
natural preference for the middle of that range, because it gives - 
the strongest resonance and the most forcible consonant. Hence 
the resonances both of s and f appear to be far less mobile, in 
English at least, than the resonances previously studied. They are 
drawn aside like the others, no doubt, by adjacent phones, but how 
far and how quickly can only be ascertained from a study of con- 
nected phonograms. 
When it is endeavoured to determine for s and f respectively 
this neutral type, a curious difference reveals itself — the accidental, 
difference already referred to ; the neutral pitch of f is some 4 or 5 
semitones lower than that of s. This is heard at once when s and 
f are articulated in immediate succession, in either order ; and a 
careful comparison of the articulations discloses the reason. It is 
evident that the upward arching of the fore-tongue, which takes 
place in passing from s to f (see ante) disposes the organs to the 
formation of a longer palatal passage ; whilst in s the necessity of 
avoiding friction in that passage causes the tongue to hold off 
somewhat from the palate, and disposes it to trend more rapidly 
downwards into the open pharynx. Thus it happens that the 
most convenient f passage is about \ longer than the most con- 
