imaginative faculty. Both may equally intend to speak the truth. The
information of the first is simply defective. That of the second is much
more dangerous. The first gives, in answer to a question asked about a
thing that has been before his eyes perhaps for years, information
exceedingly imperfect, or says, he does not know. He has never observed.
And people simply think him stupid.

The second has observed just as little, but imagination immediately
steps in, and he describes the whole thing from imagination merely,
being perfectly convinced all the while that he has seen or heard it; or
he will repeat a whole conversation, as if it were information which had
been addressed to him; whereas it is merely what he has himself said to
somebody else. This is the commonest of all. These people do not even
observe that they have _not_ observed, nor remember that they have
forgotten.

Courts of justice seem to think that anybody can speak "the whole truth,
and nothing but the truth," if he does but intend it. It requires many
faculties combined of observation and memory to speak "the whole truth,"
and to say "nothing but the truth."

"I knows I fibs dreadful; but believe me, Miss, I never finds out I have
fibbed until they tells me so," was a remark actually made. It is also
one of much more extended application than most people have the least
idea of.

Concurrence of testimony, which is so often adduced as final proof, may
prove nothing more, as is well known to those accustomed to deal with
the unobservant imaginative, than that one person has told his story a
great many times.

I have heard thirteen persons "concur" in declaring that fourteenth, who
had never left his bed, went to a distant chapel every morning at seven
o'clock.

I have heard persons in perfect good faith declare, that a man came to
dine every day at the house where they lived, who had never dined there
once; that a person had never taken the sacrament, by whose side they
had twice at least knelt at Communion; that but one meal a day came out
of a hospital kitchen, which for six weeks they had seen provide from
three to five and six meals a day. Such instances might be multiplied
_ad infinitum_ if necessary.

[2]
This is important, because on this depends what the remedy will be. If a
patient sleeps two or three hours early in the night, and then does not
sleep again at all, ten to one it is not a narcotic he wants, but food
or stimulus, or perh