ng a thunderstorm saved one from being struck
by the thunner bolt." In North and South America the bull roarer, on the
other hand, is used, not to avert, but magically to produce thunder and
lightning. {91} Among the Kaitish thunder is caused by the churinga of
their "sky dweller," Atnatu.

Wherever the toy is used for a superstitious purpose, it is, so far,
_churinga_, and, so far, modern Aberdeenshire had the same _churinga
irula_ as the Arunta. The object was familiar to palaeolithic man.

XXIV--CONCLUSION OF ARGUMENT FROM SURVIVALS IN MAGIC

I have made it perfectly certain that magic stones, "witch stones,"
"charm stones," and that _churinga irula_, wooden magical slats of wood,
exist in Australia and other savage regions, and survive, as magical,
into modern British life. The point is beyond doubt, and it is beyond
doubt that, in many regions, the stones, and the slats of wood, may be
inscribed with archaic markings, or may be uninscribed. This will be
proved more fully later. Thus Pictish, like modern British civilisation,
may assuredly have been familiar with charm stones. There is no _a
priori_ objection as to the possibility.

Why should Pictish stones _not_ be inscribed with archaic patterns
familiar to the dwellers among inscribed rocks, perhaps themselves the
inscribers of the rocks? Manifestly there is no _a priori_
improbability. I have seen the archaic patterns of concentric circles
and fish spines, (or whatever we call the medial line with slanting side
lines,) neatly designed in white on the flag stones in front of cottage
doors in Galloway. The cottagers dwelt near the rocks with similar
patterns on the estate of Monreith, but are not likely to have copied
them; the patterns, I presume, were mere survivals in tradition.

The Picts, or whoever they were, might assuredly use charm stones, and
the only objection to the idea that they might engrave archaic patterns
on them is the absence of record of similarly inscribed small stones in
Britain. The custom of using magic stones was not at all incongruous
with the early Pictish civilisation, which retained a form of the Family
now long outworn by the civilisation of the Arunta. The sole objection
is that _a silentio_, silence of archaeological records as to _inscribed_
small stones. That is not a closer of discussion, nor is the silence
absolute, as I shall show.

Moreover, the appearance of an unique and previously unheard-of set