Science would ask for their production; and even if people had guessed at
this, it is human to lose or destroy old papers.

The remoteness of the occurrences is more remarkable, for, if these things
happen, why were so few recent cases discovered? Again, the seers were
sometimes under anxiety, though such cases were excluded from the final
computation: they frequently knew that the person seen was in bad health:
they were often very familiar with his personal aspect. Now what are
called 'subjective hallucinations,' non-coincidental hallucinations,
usually represent persons very familiar to us, persons much in our minds.
I know seven cases in which such hallucinations occurred. 1, 2, of husband
to wife; 3, son to mother; 4, brother to sister; 5, sister to sister;
6, cousin (living in the same house) to cousin; 7, friend (living a mile
away) to two friends. In no case was there a death-coincidence. Only in
case 4 was there any kind of coincidence, the brother having intended to
do (unknown to the sister) what he was seen doing--driving in a dog-cart
with a lady. But he had _not_ driven. We cannot, of course, _prove_ that
these seven cases were _not_ telepathic, but there is no proof that they
were. Now most of the coincidental cases, on which the Committee relied as
their choicest examples, represented persons familiarly known to the
seers. This looks as if they were casual; but, of course, if telepathy
does exist, it is most likely (as Hegel says) to exist between kinsfolk
and friends.[16]

The dates might be fresher!

In case 1, percipient knew that his aunt in England (he being in
Australia) was not very well. No anxiety.

2. Casual acquaintance. No anxiety. Case of accident or suicide.