inscribed in the scarf at the moment when he last wore it.

In this particular case, considering that all relations with the
living were definitely and undeniably severed, I can see no other
explanations beyond these two. They are both equally astounding
and land us suddenly in a world of fable and enchantment which we
thought that we had left for good and all. If we do not adopt the
theory of the tell-tale scarf, we must accept that of the
spiritualists, who maintain that the spirits communicate with us
freely. It is possible that they may find a serious argument in
this case. But a solitary fact is not enough to support a theory,
all the more so as the one in question will never be absolutely
safe from the objection that could be raised if the case were one
of murder, which is possible, after all, and cannot be actually
disproved. We must, therefore, while awaiting other similar and
more decisive facts, if any such are conceivable, return to those
which are, so to speak, laboratory facts, facts which are only
denied by those who will not take the trouble to verify them; and
to interpret these facts there are only the two theories which we
mentioned above, before this digression; for, in these cases,
which are unlike those of automatic speech or writing, we have
not as a rule to consider the possibility of any intervention of
the dead. As a matter of fact, the best-known psychometers are
very rarely spiritualists and claim no connection with the
spirits. They care but little, as a rule, about the source of
their intuitions and seem very little interested in their exact
working and origin. Now it would be exceedingly surprising if,
acting and speaking in the name of the departed, they should be
so consistently ignorant of the existence of those who inspire
them; and more surprising still if the dead, whom in other