y are
specially likely to fall into this kind of error. Not only so: when they
grow up and their early recollections lose their definiteness, becoming
a few fragments saved from a lost past, it must pretty certainly happen
that if any ideas derived from these recitals are preserved, they will
simulate the form of memories. Thus, I have often caught myself for a
moment under the sway of the illusion that I actually visited the
Exhibition of 1851, the reason being that I am able to recall the
descriptions given to me of it by my friends, and the excitement
attending their journey to London on the occasion. It is to be added
that repetition of the act of imagination will tend still further to
deepen the subsequent feeling that we are recollecting something. As
Hartley well observes, a man, by repeating a story, easily comes to
suppose that he remembers it.[130]

Here, then, we have another source of error that we must take into
account in judging of the authenticity of an autobiographical narration
of the events of childhood. The more imaginative the writer, the greater
the risk of illusion from this source as well as from that of
dream-fancies. It is highly probable, indeed, that in such full and
explicit records of very early life as those given by Rousseau, by
Goethe, or by De Quincey, some part of the quasi-narrative is based on
mental images which come floating down the stream of time, not from the
substantial world of the writer's personal experience, but from the airy
region of dream-land or of waking fancy.

It is to be added that even when the quasi-recollection does answer to a
real event of childish history, it may still be an illusion. The fact
that others, in narrating events to us, are able to awaken imaginations
that afterwards appear as past realities, suggests that much of our
supposed early recollection owes its existence to what our parents and
friends have from time to time told us respecting the first stages of
our history.[131] We see, then, how much uncertainty attaches to all
autobiographical description of very early life.

Modern science suggests another possible source of these distinct
spectra of memory. May it not happen that, by the law of hereditary
transmission, which is now being applied to mental as well as bodily
phenomena, ancestral experiences will now and then reflect themselves
in our mental life, and so give rise to apparently personal
recollections? No one can say that this is no