he rest of the civilized world, for the Arabic,--both greater changes
than that now proposed. A change of numeration is truly a more serious
matter, yet the difficulty may not be as great as our apprehensions
paint it. Its inauguration must not be compared with that of French
gradation, which, though theoretically perfect, is practically absurd.

Decimal numeration grew out of the fact that each person has ten fingers
and thumbs, without reference to science, art, or commerce. Ultimately
scientific men discovered that it was not the best for certain purposes,
consequently that a change might be desirable; but as they were not
disposed to accommodate themselves to popular practices, which they
erroneously viewed, not as necessary consequences, but simply as bad
habits, they suggested a base with reference not so much to commerce as
to science. The suggestion was never acted on, however; indeed, it would
have been in vain, as Delambre remarks, for the French commission to
have made the attempt, not only for the reason he presents, but also
because it does not agree with natural division, and is therefore not
suited to commerce; neither is it suited to the average capacity of
mankind for numbers; for, though some may be able to use duodecimal
numeration and notation with ease, the great majority find themselves
equal to decimal only, and some come short even of that, except in its
simplest use. Theoretically, twelve should be preferred to ten, because
it agrees with circle measure at least, and ten agrees with nothing;
besides, it affords a more comprehensive notation, and is divisible by
6, 4, 3, and 2 without a fraction, qualities that are theoretically
valuable.

At first sight, the universal use of decimal numeration seems to be an
argument in its favor. It appears as though Nature had pointed directly
to it, on account of some peculiar fitness. It is assumed, indeed, that
this is the case, and habit confirms the assumption; yet, when
reflection has overcome habit, it will be seen that its adoption was due
to accident alone,--that it took place before any attention was paid to
a general system, in short, without reflection,--and that its supposed
perfection is a mere delusion; for, as a member of such a system, it
presents disagreements on every hand; as has been said, it has no
agreement with anything, unless it be allowable to say that it agrees
with the Arabic mode of notation. This kind of agreement it has, in
com