Now, by an effort, let this tendency to re-open be counteracted. Let the remaining current of breath be cut short. We have, then, only this, viz., so much of the syllable _āp_ as can be formed by the _closure_ of the lips. All that portion of it that is caused by their re-opening is deficient. The resulting sound seems truncated, cut short, or incomplete. It is the sound of p, _minus_ the remnant of breath. All of the sound p that is now left is formed, not by the _escape_ of the breath, but by the _arrest_ of it. The p in āp is a _final_ sound. With initial sounds the case is different. Let the lips be _closed_, and let an attempt be made to form the syllable pa by suddenly opening them. The sound appears incomplete; but its incompleteness is at the _beginning_ of the sound, and not at the end of it. In the natural course of things there would have been a current of breath _preceding_, and this current would have given a vibration, now wanting. All the sound that is formed here is formed, not by the _arrest_ of breath, but by the _escape_ of it. I feel that this account of the mechanism of the apparently simple sound p, labours under all the difficulties that attend the _description_ of a sound; and for this reason I again request the reader to satisfy himself either of its truth or of its inaccuracy, before he proceeds to the conclusions that will be drawn from it. The account, however, being recognized, we have in the sound of p, two elements:-- 1. That formed by the current of air and the closure of the lips, as in ap. This may be called the sound of breath _arrested_. 2. That formed by the current of air, and the opening of the lips, as in pa. This may be called the sound of breath _escaping_. Now what may be said of p may be said of all the other consonants, the words _tongue_, _teeth_, &c., being used instead of _lips_, according to the case. Let the sound of breath _arrested_ be expressed by π, and that of breath _escaping_ be expressed by ϖ, the two together form p (π + ϖ = p). Thus ap (as quoted above) is p - ϖ, or π; whilst pa (sounded similarly) is p - π, or ϖ. In the formation of syllables, I consider that the sound of breath arrested belongs to the first, and the sound of breath escaping to the second syllable; that if each sound were expressed by a separate sign, the word _happy_ would be divided thus, _haπ-ϖy_; and that such would be the case with all consonants between two syllables. The _whole_ consonant belongs neither to one syllable nor the other. Half of it belongs to each. The reduplication of the p in _happy_, the t in _pitted_, &c., is a mere point of spelling. * * * * * CHAPTER VI. ON QUANTITY.