of Mathonwy king of Gwyned, and his sister's son, Gwydion ab Dôn, to whom as his successor he duly taught his magic; then come the arch-enchanter Arawn, king of Annwn, and Caswallon ab Beli, represented as winning his kingdom by the sheer force of magic. To these might be added other members of the kingly families whose story shows them playing the rôle of magicians, such as Rhiannon, who by her magic arts foiled her powerful suitor, Gwawl ab Clûd, and secured as her consort the man of her choice, Pwyll prince of Dyfed. Here also, perhaps, one might mention Manawyđan ab ILyr, who, as Manannán mac Lir, figures in the stories of the Goidels of Erin and Man as a consummate wizard and first king of the Manx people: see p. 314 above. In the Mabinogi, however, no act of magic is ascribed to Manawydan, though he is represented successfully checkmating the most formidable wizard arrayed against him and his friends, to wit, ILwyd ab Kilcoed. Not only does one get the impression that the ruling class in these stories of the Welsh Goidels had their magic handed down from generation to genera. tion according to a fixed rule of maternal succession (pp. 326, 503, 505), but it supplies the complete answer to and full explanation of questions as to the meaning of the terms already mentioned, Tuatha Dé ocus Ande, and Lucht Cumachtai, together with its antithesis. Within the magic-wielding class exercising dominion over the shepherds and tillers of the soil of the country, it is but natural to suppose that the first king was the first magician or greatest medicine-man, as in the case of Manannán in the Isle of Man. This must of course be understood to apply to the early history of the Goidelíc race, or, perhaps more correctly speaking, to one of the races which had contributed to its composition: to the aborigines, let us say, by whatsoever name or names

you may choose to call them, whether Picts or Ivernians. It is significant, among other things, that our traditions should connect the potency of ancient wizardry with descent in the female line of succession, and, in any case, one cannot be wrong in assuming magic to have begun very low down in the scale of social progress, probably lower than religion, with which it is essentially in antagonism. As the crude and infantile pack of notions, collectively termed sympathetic magic-beginning with the belief that any effect may be produced by imitating the action of the cause of it, or even doing anything that would recall it 1-grew into the panoply of the magician, he came to regard himself, and to be regarded by others, as able for his own benefit and that of his friends to coerce all possible opponents, whether men or demons, heroes or gods. This left no room for the attitude of prayer and worship: religion in that sense could only come later.

1 See Frazer's Golden Bough, i. 9, where a few most instructive instances are given.

CHAPTER XII

RACE IN FOLKLORE AND MYTH

The method of philological mythology is thus discredited by the disputes of its adherents. The system may be called orthodox, but it is an orthodoxy which alters with every new scholar who enters the sacred enclosure.-ANDREW LANG.

It has been well said, that while it is not science to know the contents of myths, it is science to know why the human race has produced them. It is not my intention to trace minutely the history of that science, but I may hazard the remark, that she could not be said to have reached years of discretion till she began to compare one thing with another; and even when mythology had become comparative mythology, her horizon remained till within recent years comparatively narrow. In other words, the comparisons were wont to be very circumscribed: you might, one was told, compare the myths of Greeks and Teutons and Hindus, because those nations were considered to be of the same stock; but even within that range comparisons were scarcely contemplated, except in the case of myths enshrined in the most classical literatures of those nations. This kind of mythology was eclectic rather than comparative, and it was apt to regard myths as a mere disease of language. By-and-by, however, the student showed a preference for a larger field and a wider range; and in so doing he was, whether

consciously or unconsciously, beginning to keep step with a larger movement extending to the march of all the kindred sciences, and especially that of language.

At one time the student of language was satisfied with mummified speech, wrapped up, as it were, in the musty coils of the records of the past: in fact, he often became a mere researcher of the dead letter of language, instead of a careful observer of the breath of life animating her frame. So long as that remained the case, glottology deserved the whole irony of Voltaire's well-known account of etymology as being in fact, 'une science où les voyelles ne font rien, et les consonnes fort peu de chose.' In the course, however, of recent years a great change has come over the scene: not only have the laws of the Aryan consonants gained greatly in precision, but those of the Aryan vowels have at last been discovered to a considerable extent. The result for me and others who learnt that the Aryan peasant of idyllic habits harped eternally on the three notes of a, i, u, is that we have to unlearn this and. a great deal more: in fact, the vowels prove to be far more troublesome than the consonants. But difficult as these lessons are, the glottologist must learn them, unless he is content to remain with the stragglers who happen to be unable to move on. Now the change to which I allude, in connexion with the study of language, has been inseparably accompanied with the paying of increased attention to actual speech, with a more careful scrutiny of dialects, even obscure dialects such as the literary man is wont to regard with scorn.

Similarly the student of mythology now seeks the wherewithal of his comparisons from the mouth of the traveller and the missionary, wherever they may roam; not from the Rig-Veda or the Iliad alone, but from the rude stories of the peasant, and the wild fancies of

the savage from Tierra del Fuego to Greenland's icy mountains. The parallel may be drawn still closer. Just as the glottologist, fearing lest the written letter may have slurred over or hidden away important peculiarities of ancient speech, resorts for a corrective to the actuality of modern Aryan, so the mythologist, apt to suspect the testimony of the highly respectable bards of the Rig-Veda, may on occasion give ear to the fresh evidence of a savage, however inconsequent it may sound. The movements to which I allude in glottology and mythology began so recently that their history has not yet been written. Suffice it to say that in glottology, or the science of language, the names most intimately connected with the new departure are those of Ascoli, J. Schmidt, and Fick, those of Leskien, Brugmann, Osthoff, and De Saussure; while of the names of the teachers of the anthropological method of studying myths, several are by this time household words in this country. But, so far as I know, the first to give a systematic exposition of the subject was Professor Tylor, in his work on Primitive Culture, published first in 1871.

Such has been the intimate connexion between mythology and glottology that I may be pardoned for going back again to the latter. It is applicable in its method to all languages, but, as a matter of fact, it came into being in the domain of Aryan philology, so that it has been all along principally the science of comparing the Aryan languages with one another. It began with Sir William Jones' discovery of the kinship of Sanskrit with Greek and Latin, and for a long time it took the lead of the more closely related sciences: this proved partly beneficial and partly the reverse. In the case of ethnology, for instance, the influence of glottology has probably done more harm than good, since it has opened up a wide field for confounding race with