Everybody is familiar with the fact that the “transmission of sound,” the pulse which strikes upon the ear to produce the sensation, _is affected_ by currents of air--the direction, force, and velocity of the wind--between the initial pulse and the hearer. _How_? and _how much_? _directly_ or _indirectly_? are questions distinct from the fact itself. The distance through which guns are heard, as well as the loudness of their report, varies with the _direction_, _force_, and _velocity_ of the wind; and, in very still air, with the _aim_ of the gun itself, the _direction of the initial pulse_. For short distances, these differences may be so minute as to escape notice; just as the false proportions of a miniature picture are unobserved until the magnifier displays them. And for longer ranges, they are so small, in contrast with the magnitudes compared, as to seem rather like _accidental_ than _legitimate_ differences. But the difference is not the less real because the reality is less. Words spoken in a faint whisper are clearly heard by a listener immediately _before_ the speaker, when quite inaudible or indistinct to one at an equal distance _behind_ him. The _actual velocities_ of wind and sound differ so widely that the small fraction by which their _relative velocity_ is denoted is held as proof that the propagation of sound--_the pulse_--through distances of a few yards or feet, is not affected by currents of air: that there are no _differences in the “velocity of sound.”_ Yet the ear detects them as one of the small differences between discord and harmony in music; distinctness and confusion of speech. In music these differences may be blended by the prolonged intonation of _vowel sounds_; but in speech, whose distinct significance is due to _consonants_, “which cannot be sounded without the aid of a vowel,” these differences are fatally evident. The sharp edges of the vocal pulses, which give shape and meaning to vowel sounds, are destroyed alike by a husky voice and a puff of air. What remains is _vox et præterea nihil_. It seems to us that some of the many failures in practical acoustics come from considering the air--the material involved--as perfectly elastic. From this it is inferred that sound is not affected by the direction of the initial _pulse_: that the direction and velocity of the _effective_ pulse are not varied by currents and blasts of air. In short, that the slight inaccuracy of these assumptions will be the actual measurement of resultant error.