stimulus.

#After-oscillation.#--When the sensitive surface is subjected to the
continued action of light, the E.M. effect attains a maximum at which
it remains constant for some time. If the exposure be maintained after
this for a longer period, there will be a decline, as we found to be the
case in other instances of continued stimulation. The appearance of this
decline, and its rapidity, depends on the particular condition of the
substance.

When the sensitive element is considerably strained by the action of
light, and if that light be now cut off, there is a rebound towards
recovery and a subsequent after-oscillation. That is to say, the curve
of recovery falls below the zero point, and then slowly oscillates back
to the position of equilibrium. We have already seen an instance of this
in fig. 102. Above is given a series of records showing the appearance
of decline, from too long-continued exposure and recovery, followed by
after-oscillation on the cessation of light (fig. 105). Certain visual
analogues to this phenomenon will be noticed later.

[Illustration: FIG. 105.--AFTER-OSCILLATION
Exposure of one minute followed by obscurity of one minute. Note the
decline during illumination, and after-oscillation in darkness.]

#Abnormal effects.#--We have already treated of all the normal effects of
the stimulus of light on the retina, and their counterparts in the
sensitive cell. But the retina undergoes molecular changes when injured,
stale, or in a dying condition, and under these circumstances various
complicated modifications are observed in the response.

[Illustration: FIG. 106.--TRANSIENT POSITIVE AUGMENTATION GIVEN BY THE
FROG'S RETINA ON THE CESSATION OF LIGHT L (WALLER)]

#1. Preliminary negative twitch.#--When the light is incident on the
frog's retina, there is sometimes a transitory negative variation,
followed by the normal positive response. This is frequently observed in
the sensitive cell (see fig. 96, _b_).

#2. Reversal of response.#--Again, in a stale retina, owing to molecular
modification the response is apt to undergo reversal (Waller). That is
to say, it now becomes negative. In working with the same sensitive cell
on different days I have found it occasionally exhibiting this reversed
response.

#3. Transient rise of current on cessation of light.#--Another very
curious fact observed in the retina by Kuhne and Steiner is that
immediately on the stoppage of light there is somet