charm and many idols. I doubt if even a single run of luck be
necessary. I am sure that if an elder boy said that 'the pretty fish
was lucky--of course it was,' all the lesser boys would believe it,
and in a week it would be an accepted idol. And I suspect the Nestor
of a savage tribe--the aged repository of guiding experience--would
have an equal power of creating superstitions. But if once created
they are most difficult to eradicate. If any one said that the
amulet was of certain efficacy--that it always acted whenever it was
applied--it would of course be very easy to disprove; but no one
ever said that the 'pretty fish' always brought luck; it was only
said that it did so on the whole, and that if you had it you were
more likely to be lucky than if you were without it. But it requires
a long table of statistics of the results of games to disprove this
thoroughly; and by the time people can make tables they are already
above such beliefs, and do not need to have them disproved. Nor in
many cases where omens or amulets are used would such tables be easy
to make, for the data could not be found; and a rash attempt to
subdue the superstition by a striking instance may easily end in
confirming it. Francis Newman, in the remarkable narrative of his
experience as a missionary in Asia, gives a curious example of this.
As he was setting out on a distant and somewhat hazardous
expedition, his native servants tied round the neck of the mule a
small bag supposed to be of preventive and mystic virtue. As the
place was crowded and a whole townspeople looking on, Mr. Newman
thought that he would take an opportunity of disproving the
superstition. So he made a long speech of explanation in his best
Arabic, and cut off the bag, to the horror of all about him. But as
ill-fortune would have it, the mule had not got thirty yards up the
street before she put her foot into a hole and broke her leg; upon
which all the natives were confirmed in their former faith in the