order of beings
with supernatural powers referred to several times in the foregoing
pages: I mean Witches. I adduced in Chapter X. a Tirolese tale, a
variant of the Melusina type, wherein the wife was a witch. It will have
been obvious to every reader that the tale is simply that of Cupid and
Psyche with the parts reversed; and I might urge that Cupid and the
witch were beings of precisely the same nature. Waiving this for the
moment, however, no one will deny that the witch takes the place of the
Swan-maiden, or fairy, in other stories of the group. But perhaps it may
be suggested that the name _witch_ (_Angana_, _Hexe_) has got into the
story by accident; and that not a witch in our sense of the word, but a
ghost from the dead, is really meant. There might be something to be
said for this if there were any substantial distinction to be made
between ghosts and witches and fairies. In the tales and superstitions
discussed in the present volume we have found no distinction. Whether it
be child-stealing, transformation, midnight meetings, possession and
gift of enchanted objects, spell-binding, or whatever function, or
habit, or power be predicted of one, it will be found to be common to
the three. I conclude, therefore, that they are all three of the same
nature. This is what a consideration of the superstitions of savages
would lead me to expect. The belief in fairies, ghosts, and witches is a
survival of those superstitions. It is, of course, not found in equal
coherence, equal strength of all its parts, equal logic (if I may so
express it) everywhere. We must not be surprised if, as it is gradually
penetrated by the growing forces of civilization, it becomes
fragmentary, and the attributes of these various orders of supernatural
beings begin to be differentiated. They are never completely so; and the
proof of this is that what is at one place, at one time, or by one
people, ascribed to one order, is at another place, at another time, or
by another people, ascribed to another order. The nature of the
classical deities was identical too; and hence Cupid and the witch of
the Tirolese tale are the masculine and feminine counterparts of the
same conception.

Lastly, a few words must be expended on a totally different theory
lately put forward by Mr. MacRitchie. This theory is not altogether a
new one; it has been before the world for many years. But Mr. MacRitchie
has, first in "The Archaeological Review," and since then m