pens that while they mainly coincide, they, so to speak,
shingle over each other at random, and it follows: 1. That the Number of
Sounds in different Languages is not uniform; 2. That of any two
Languages compared, one will chance to have several sounds not heard in
the other; and, 3. The erroneous impression is made upon the casual and
superficial observer that in the aggregate of all Languages there must
be an immense number of sounds; whereas, in fact, the total Alphabet of
Vocal Sounds in nature, like the Gamut of Colors or Musical Tones, is
quite limited, if we attend only to those which distinctly differ, or
stand at appropriate and appreciable distances from each other.

Further to illustrate: Assume that there are, capable of being clearly
discriminated by the human ear, say sixty-four or seventy-two distinct
Elementary Sounds of the human voice, in all--as many, for example, as
there are Chemical Elements; some existing Languages select and make use
of twenty, some of twenty-four, some of thirty, and some of forty of
these sounds, omitting the rest.

But--and here is a very important point and a real discovery in this
investigation--it will be found, if closely attended to, that a certain
selection of one half of this number, say thirty-two or thirty-six of
these sounds, embraces the whole body of vocal elements _usually
occurring_ in all the forms of speech on the planet; the remaining half
consisting of rare, exceptional, and, we may nearly say, useless sounds.
This statement will again be better understood by analogy with what
regards the Elements of Chemistry. Just about one half of the known
elements of matter occur with frequency, and enter into useful and
ordinary combinations to produce the great mass of known substances. The
remaining half are unfrequent, obscure, and relatively unimportant; some
of them never having been seen even by many of our most eminent
chemists. Even should a few new elements be discovered, it cannot be
anticipated that any one of them should prove to be of leading
importance, like oxygen, carbon, or sulphur.

On the other hand, should some future great chemical discovery realize
the dream of the alchemists, and enable us to transmute iron into gold,
and indeed every chemical Element into every other chemical Element
(convertible identity), still the sixty-four (nearly) Chemical Elements
now known would remain the real Elements of Organic and Inorganic
Compounds, in a sense ju