231 
1 . 897 - 98 .] Mr E. J. Lloyd on Consonant- Sounds. 
no longer quite neutral h to the ear, when it is drawn much away 
from this unbiassed articulation. And the reason is clear. This 
neutral h keeps the upper and lower sides of the voice-passage as 
far as possible, parallel and equidistant. We cannot depart very 
far from that arrangement in any direction, without making the 
passage distinctly narrower in some places than in others ; and if 
narrower, then more obstructive and more frictional. Whenever 
the friction of h is thus localised, it at once departs more or less 
from the simple neutral sound of h. Yet this is the kind of h 
which usually occurs in actual speech. Whether the departure 
from type can he recognised by the hearer or not, depends on the 
extent of the departure, and the acuteness of the listening ear. 
One kind of departure is patent to everybody : it is when the con- 
striction is so local and so tight as to transform the h into a x or 
Precise and forcible speakers may often he heard to pronounce lie 
who as xi yu. But A. J. Ellis, the translator of Helmholtz, 
noticed many years ago that a mere ordinary li often carries the 
impress of the following vowel. In other words, if the pronuncia- 
tion of a word beginning with h is arrested before the vowel is 
actually begun, the ear can often recognise, from the h alone, what 
vowel was going to follow. 
This phenomenon is only one instance of a habit of language so 
wide-spread that it may he called a law. It is simply that the 
organs are always endeavouring during the articulation of any 
phone, so far as the conditions of that articulation will allow, to 
put themselves into the readiest position for commencing the 
articulation of the next phone. Conceive in your mind the in- 
tention of pronouncing the word blue : you will find your tongue 
in the l position before you have even begun to sound the b. In 
the phenomenon called umlaut , so important to all the Germanic 
languages, this influence extends even further hack ; for it modifies 
the preceding vowel. The reason why our h conveys, as a rule, 
not only the sound of h, hut an inference of the following vowel, 
is simply that the organs have already, during the h , assumed 
something like the prospective vowel-position; so much so, indeed, 
that the oral or forward resonance of the vowel is produced and 
heard, with the distinctive breathy timbre of the h. Thus a quick 
ear infers the vowel, though it does not really hear it ; for the ap- 
