230 Proceedings of Boyal Society of Edinburgh. [sess. 
studied. In h , as in each consonant treated hitherto, our first aim 
must be to define the two variable conditions — those which serve 
to differentiate it from every fricative previously described. 
These are (1) the friction, and (2) the resonance. It is the dis- 
tinguishing feature of the h friction that it does not take place at 
any definite point : in fact it is convenient to regard as the purest 
or most typical h that kind of aspirate which creates friction at all 
points of the voice-passage, from the pharynx to the external air. 
This implies that in such an h the degree of obstruction is fairly 
uniform ; and this in turn implies that the resonant passage is of 
fairly uniform calibre throughout. This position of the organs 
does not materially differ from that which they naturally assume 
in sighing ; and thus the pitch of a sigh differs little from that of 
this typical h. In me it seems to he about e 3 1320 v.d. 
Leaving h for a moment, it may he remarked that not only h, 
hut every fricative, tends to what may he called a neutral type, 
when pronounced in isolation. Underlying all our articulations, 
there is the principle of economy of adjustment. When we start 
from a state of rest and end in a state of rest, the articulation de- 
parts usually from the position of rest, the ordinary silent position, 
so far as is necessary, and no further. Hence f 0, and no less 
than h, have what may he called a neutral pitch in every in- 
dividual. In myself I find that this neutral pitch is, for /, about 
f, 1408 v.d. ; for x , f% 1491 v.d. ; and for 6, g * 1584 v.d. This 
steady rise from h to 0 corresponds well with the changes in the 
oral passage. In k the passage embraces the whole distance from 
the pharynx to the external air, and it is wide all through ; in / 
it is slightly foreshortened, and much narrowed at one end : in x 
it recovers a little length forward, hut loses considerably in length 
behind ; and in 6 foreshortening again occurs in stronger form, 
than in f; and the advance of the tongue- tip draws the whole 
tongue forward, and causes the pharyngeal cavity to extend itself 
part way up the hack of the tongue. 
But in h this neutral or isolated value has some special features. 
It is true that the articulation and pitch of h, even more than that 
of /, 6, x, or Xj is pulled hither and thither, in actual speech, by 
the adjacent phones. But, through all these changes, / is still /, 
6 is 6, and nothing more : x and x are still either x or x : but h i s 
