of the Forms that appear here and
there, are dominoes, cards, counters, an abacus, the fingers,
counting by coins, feet and inches (a yellow carpenter's rule
appears in one case with 56 in large figures upon it), the country
surrounding the child's home, with its hills and dales, objects in
the garden (one scientific man sees the old garden walk and the
numeral 7 at a tub sunk in the ground where his father filled his
watering-pot). Some associations seem connected with the objects
spoken of in the doggerel verses by which children are often taught
their numbers.

But the paramount influence proceeds from the names of the numerals.
Our nomenclature is perfectly barbarous, and that of other civilised
nations is not better than ours, and frequently worse, as the French
"quatre-vingt dix-huit," or "four score, ten and eight," instead of
ninety-eight. We speak of ten, eleven, twelve, thirteen, etc., in
defiance of the beautiful system of decimal notation in which we
write those numbers. What we see is one-naught, one-one, one-two, etc.,
and we should pronounce on that principle, with this proviso, that
the word for the "one" having to show both the place and the value,
should have a sound suggestive of "one" but not identical with it.
Let us suppose it to be the letter _o_ pronounced short as in
"on," then instead of ten, eleven, twelve, thirteen, etc., we might
say _on-naught, on-one, on-two, on-three_, etc.

The conflict between the two systems creates a perplexity, to which
conclusive testimony is borne by these numerical forms. In most of
them there is a marked hitch at the 12, and this repeats itself at
the 120. The run of the lines between 1 and 20 is rarely analogous
to that between 20 and 100, where it usually first becomes regular.
The 'teens frequently occupy a larger space than their due. It is not
easy to define in words the variety of traces of the difficulty and
annoyance caused by our unscientific nomenclature, that are
portrayed vividly, and, so to speak, painfully in these pictures.
They are indelible scars that testify to the effort and ingenuity
with which a sort of compromise was struggled for and has finally
been effected between the verbal and decimal systems. I am sure that
this difficulty is more serious and abiding than has been suspected,
not only from the persistency of these twists, which would have long
since been smoothed away if they did not continue to subserve some
useful purpose, but als