g before he could speak a
word. But this did not go so far as the framing of syllables. It is not
easy in this to trace so clearly the framing of a concept as attaching
itself directly to onomatopoetic forms as it is in a case communicated
by Romanes. A child that was beginning to talk, saw and heard a duck on
the water, and said _quack_. Thereafter the child called, on the one
hand, all birds and insects, on the other hand, all liquids, _quack_.
Finally, it called all coins also _quack_, after having seen an eagle
on a French sou. Thus the child came, by gradual generalization, to the
point of designating a fly, wine, and a piece of money by the same
onomatopoetic word, although only the first perception contained the
characteristic that gave the name.

Another case is reported by Eduard Schulte: A boy of a year and three
quarters applied the joyous outcry _ei_ (which may be an imitated
interjection), modifying it first into _eiz_, into _aze_, and then into
_ass_, to his wooden goat on wheels, and covered with rough hide; _eiz_,
then, became exclusively a cry of joy; _ass_, the name for everything
that moved along--e. g., for animals and his own sister and the wagon;
also for everything that moved at all; finally, for everything that had
a rough surface. Now, as this child already called all coverings of the
head and covers of cans _huta_, when he saw, for the first time, a fur
cap, he at once christened it _ass-huta_. Here took place a decided
subordination of one concept to another, and therewith a new formation
of a word. How broad the comprehensiveness of the concept designated
_huta_ was, is perceived especially in this, that it was used to express
the wish to have objects at which the child pointed. He liked to put all
sorts of things that pleased him upon his head, calling them _huta_. Out
of the _huta_, for "I should like to have that as a hat" grew, then,
after frequent repetition, "I should like that." There was in this case
an extension of the narrower concept, after it had itself experienced
previously a differentiation, and so a limitation, by means of the
suffix _ass_. These examples show how independent of words the formation
of concepts is. With the smallest stock of words the concepts are yet
manifold, and are designated by the same word when there is a lack of
words for the composition of new words, and so for fresh
word-formation.

The formation of words out of interjections without imitation has no