, we must prepare the
"environment" in a definite manner, and from this environment offer
the child the external means directly necessary for him.

This is the _positive_ fact which my experiment has rendered concrete.
Hitherto the liberty of the child has been vaguely discussed; no
clearly defined limit has been established between liberty and
abandonment. We were told: "Liberty has its limits," "Liberty must be
properly understood." But a special method indicating "how liberty
should be interpreted, and what is the intuitive _quid_ which ought
to co-exist with it," had not been determined.

The establishment of such a method should open up a new path to all
education.

* * * * *

It is therefore necessary that the environment should contain the
means of auto-education. These means cannot be "taken at random"; they
represent the result of an experimental study which cannot be
undertaken by all, because a scientific preparation is necessary for
such delicate work; besides, like all experimental study, it is
laborious, prolonged, and exact. Many years of research are required,
before the means really _necessary_ for _psychical development_ can be
set forth. Those educationalists who leave the great question of the
liberty of the pupil to the good sense or to the preparation of the
master are very far from solving the problem of liberty. The greatest
scientist, or the person most fitted by nature to teach, could never
of himself discover such, because, to preparation and natural gifts,
the further factor of _time_ must be added--the long period of
preparatory experiment. Therefore a _science_ which has already
_provided the means_ for self-education must exist beforehand. To-day,
he who speaks of liberty in the schools ought at the same time to
exhibit objects--approximating to a scientific apparatus--which will
make such liberty possible.

The scientific instrument must be constructed upon a basis of
_exactitude_. Just as the lenses of the physicist are constructed in
accordance with the laws of the refraction of light, so the pedagogic
instrument should be based on the _psychical manifestations_ of the
child.

Such an instrument may be compared to a systematized "mental test." It
is not, however, established upon a basis of external measurement, for
the purpose of estimating the amount of instantaneous psychical
reaction which it produces; it is, on the contrary, a stimulus which
is itself deter