s--are phases of savage thought,
every one of which has been incorporated in the myth of the
Swan-maidens, and every one of which, except one special and very
limited development of the doctrine of Spirits, is ignored in
Liebrecht's theory. The theory is, indeed, an admirable illustration of
the danger of reasoning without a sufficiently wide area of induction.
Liebrecht's mistake on the present occasion was twofold: he only dealt
with one or, at most, two types of the myth; and he ignored the savage
variants. Had he taken into consideration other types--such as Hasan,
the Marquis of the Sun, the Star's Daughter;--had he been aware of the
savage variants all over the world, he would not have formed a theory so
inconsistent with the facts, and so little fitted to solve the problems
propounded, not merely by the phenomena of the Swan-maiden group, but by
those of other tales in which supernatural beings intervene.

In reasoning by induction, the greater the number of facts taken into
account, the greater the probability of sound reasoning; and therefore
the greater the number of facts a theory will explain, the more likely
it is to be true. Had Liebrecht's theory touched only the Swan-maiden
group, it would have been more convenient to discuss it in the last
chapter. But inasmuch as its truth would involve much wider issues, it
seemed better to reserve it to be dealt with here. For if the theory be
valid for Melusina, the Lady of the Van Pool, and other water-nymphs, it
is valid also for the "water-woman" who, in a Transylvanian story, dwelt
in a lake in the forest between Mehburg and Reps. She had two sons,
whose father was a man, and the younger of whom became king of that
land. But when the Saxon immigration took place the incomers cut down
the wood; the lake dried up, and as it dried up, the lives of the
water-spirit and her son gradually sank lower and lower, and at last
were extinguished with the extinction of the lake.[244] Now I will
venture to say that this story is to be explained satisfactorily on no
theory yet broached, unless it be the theory that we have in it a
survival of the savage doctrine of Spirits. Least of all it is to be
explained by any adaptation of what I may call the Ghost
theory,--namely, that the water-spirit and her son were already the
spirits of dead human beings.

Leaving this one example of the value of Liebrecht's theory, as applied
to water-spirits, to stand for all, I turn to another