I look at the word as a mere collection of letters.

"Of my two daughters, one sees the colours quite differently from
this (A, blue; E, white; I, black; O, whity-brownish; U, opaque brown).
The other is only heterodox on the A and O; A being with her black,
and O white. My sister and I never agreed about these colours, and I
doubt whether my two brothers feel the chromatic force of the vowels
at all."

I give this instance partly on account of the hereditary interest. I
could add cases from at least three different families in which the
heredity is quite as strongly marked.

Fig. 69 fills the whole of the middle column of Plate IV., and
contains specimens from a large series of coloured illustrations,
accompanied by many pages of explanation from a correspondent,
Dr. James Key of Montagu, Cape Colony. The pictures will tell their
own tale sufficiently well. I need only string together a few brief
extracts from his letters, as follows:--

"I confess my inability to understand visualised numerals; it is
otherwise, however, with regard to colour associations with letters.
Ever since childhood these have been distinct and unchanging in my
consciousness; sometimes, although very seldom, I have mentioned them,
to the amazement of my teachers and the scorn of my comrades. A is
brown. I say it most dogmatically, and nothing will ever have the
effect, I am convinced, of making it appear otherwise! I can imagine
no explanation of this association. [He goes into much detail as to
conceivable reasons connected with his childish life to show that
none of these would do.] Shades of brown accompany to my mind the
various degrees of openness in pronouncing A. I have never been
destitute in all my conscious existence of a conviction that E is a
clear, cold, light-gray blue. I remember daubing in colours, when
quite a little child, the picture of a jockey, whose shirt received
a large share of E, as I said to myself while daubing it with grey.
[He thinks that the letter I may possibly be associated with black
because it contains no open space, and O with white because it does.]
The colour of R has been invariably of a copper colour, in which a
swarthy blackness seems to intervene, visually corresponding to the
trilled pronunciation of R. This same appearance exists also in J, X,
and Z."

The upper row of Fig. 69 shows the various shades of brown,
associated with different pronunciations of the letter A, as in
"fame," "can," "cha