allacies," as, for
example, by Dr. Carpenter (_Human Physiology_, ch. x.). While preferring
the term "illusion," I would not forget to acknowledge my indebtedness
to Dr. Carpenter, who first set me seriously to consider the subject of
mnemonic error.

[112] From this it would appear to follow that, so far as a percept is
representative, recollection must be re-representative.

[113] The relation of memory to recognition is very well discussed by M.
Delboeuf, in connection with a definition of memory given by
Descartes. (See the article "Le Sommeil et les Reves," in the _Revue
Philosophique_, April, 1880, p. 428, _et seq._)

[114] A very interesting account of the most recent physiological theory
of memory is to be found in a series of articles, bearing the title, "La
Memoire comme fait biologique," published in the _Revue Philosophique_,
from the pen of the editor, M. Th. Ribot. (See especially the _Revue_ of
May, 1880, pp. 516, _et seq._) M. Ribot speaks of the modification of
particular nerve-elements as "the static base" of memory, and of the
formation of nerve-connections by means of which the modified element
may be re-excited to activity as "the dynamic base of memory" (p. 535).

[115] What constitutes the difference between such a progressive and a
retrogressive movement is a point that will be considered by-and-by.

[116] It is not easy to say how far exceptional conditions may serve to
reinstate the seemingly forgotten past. Yet the experiences of dreamers
and of those who have been recalled to consciousness after partial
drowning, whatever they may prove with respect to the revivability of
remote experiences, do not lead us to imagine that the range of our
definitely localizing memory is a wide one.

[117] _Der Zeitsinn nach Versuchen_, p. 36, _et seq._

[118] _Physiologische Psychologie_, p. 782.

[119] Wundt refers these errors to variations in the state of
preadjustment of the attention to impressions and representations,
according as they succeed one another slowly or rapidly. There is little
doubt that the effects of the state of tension of the apparatus of
attention, are involved here, though I am disposed to think that Wundt
makes too much of this circumstance. (See _Physiologische Psychologie_,
pp. 782, 783. I have given a fuller account of Wundt's theory in _Mind_,
No. i.)

[120] Strictly speaking, it would occupy more time, since the effort of
recalling each successive link in the chain wou