on: and it is convenient also to
say that _son_ and _sun_ and _heir_ and _air_ are homophones without
explaining that it is meant that they are mutually homophonous, which
is evident. A physician congratulating a friend on the birth of his
first-born might say, 'Now that you have a son and heir, see that he
gets enough sun and air'.]

Homophony is between words as _significant_ sounds, but it is needful
to state that homophonous words must be _different_ words, else we
should include a whole class of words which are not true homophones.
Such words as _draft_, _train_, _board_, have each of them separate
meanings as various and distinct as some true homophones; for
instance, a draught of air, the miraculous draught of fishes, the
draught of a ship, the draft of a picture, or a draught of medicine,
or the present draft of this essay, though it may ultimately appear
medicinal, are, some of them, quite as distinct objects or notions
as, for instance, _vane_ and _vein_ are: but the ambiguity of _draft_,
however spelt, is due to its being the name of anything that is
_drawn_; and since there are many ways of drawing things, and
different things are drawn in different ways, the _same word_ has come
to carry very discrepant significations.

Though such words as these[2] are often inconveniently and even
distressingly ambiguous, they are not homophones, and are therefore
excluded from my list: they exhibit different meanings of one word,
not the same sound of different words: they are of necessity present,
I suppose, in all languages, and corresponding words in independent
languages will often develop exactly corresponding varieties of
meaning. But since the ultimate origin and derivation of a word is
sometimes uncertain, the scientific distinction cannot be strictly
enforced.

[Footnote 2: Such words have no technical class-name; they are merely
extreme examples of the ambiguity common to most words, which grows
up naturally from divergence of meaning. True homophones are separate
words which have, or have acquired, an illogical fortuitous identity.]

[Sidenote: False homophones.]

Now, wherever the same derivation of any two same-sounding words is
at all doubtful, such words are practically homophones:--and again in
cases where the derivation is certainly the same, yet, if the ultimate
meanings have so diverged that we cannot easily resolve them into one
idea, as we always can _draft_, these also may be practically recko