and it is shown not only in the folklore which has been preserved for us by The Four Masters, but also by the usual kind of story about a neglected well door. These remarks suggest the question whether it was not one of the notions which determined surface burials, that is, burials in which no cutting of the ground took place, the cists or chambers and the bodies placed in them being covered over by the heaping on of earth or stones brought from a more or less convenient distance. It might perhaps be said that all this only implied individuals of a character to desecrate the ground and call forth the displeasure of the divinities concerned; and for that suggestion folklore parallels, it is true, could be adduced. But it is hardly adequate : the facts seem to indicate a more general objection on the part of the powers in point; and they remind one rather of the clause said to be inserted in mining leases in China with the object, if one may trust the newspapers, of preventing shafts from being sunk below a certain depth, for fear of offending the susceptibilities of the demons or dragons ruling underground.

It is interesting to note the fact, that Celtic folklore connects the underground divinities intimately with water; for one may briefly say that they have access wherever water can take them. With this qualification the belief may be said to have lingered lately in Wales, for instance, in connexion with ILyn Barfog, near Aberdovey. 'It is believed to be very perilous,' Mr. Pughe says, p. 142 above, 'to let the waters out of the lake'; and not long before he wrote, in 1853, an aged inhabitant of the district informed him 'that she recollected this being done during a period of long drought, in order to procure motive power for ILyn Pair Mill, and that long-continued heavy rains followed.' Then we have the story related to Mr. Reynolds as to Lyn y

Fan Fach, how there emerged from the water a huge hairy fellow of hideous aspect, who stormed at the disturbers of his peace, and uttered the threat that unless they left him alone in his own place he would drown a whole town. Thus the power of the water spirit is represented as equal to producing excessive wet weather and destructive floods. He is in all probability not to be dissociated from the afanc in the Conwy story which has already been given (pp. 130-3). Now the local belief is that the reason why the afanc had to be dragged out of the river was that he caused floods in the river and made it impossible for people to cross on their way to market at ILanrwst. Some such a local legend has been generalized into a sort of universal flood story in the late Triad, iii. 97, as follows:-'Three masterpieces of the Isle of Prydain: the Ship of Nefyd Naf Neifion, that carried in her male and female of every kind when the Lake of ILion burst; and Hu the Mighty's Ychen Bannog dragging the afanc of the lake to land, so that the lake burst no more; and the Stones of Gwyđon Ganhebon, on which one read all the arts and sciences of the world.' A story similar to the Conwy one, but no longer to be got so complete, as far as I know, seems to have been current in various parts of the Principality, especially around ILyn Syfađon and on the banks of the Anglesey pool called yn yr Wyth Eidion, 'the Pool of the Eight Oxen,' for so many is Hu represented here as requiring in dealing with the Anglesey afanc. According to Mr. Pughe of Aberdovey, the same feat was performed at Lyn Barfog, not, however, by Hu and his oxen, but by Arthur and his horse. To be more exact the task may be here considered as done by Arthur superseding Hu: see p. 142 above. That, however, is of no consequence here, and I return to the afanc: the Fan Fach legend told to Mr. Reynolds

the family only counts once. Thus his descendants Ad and Bd in the sixth generation or lifetime, are fourth cousins separated from one another by nine ancestors: that is, they are related in the ninth âch. In other words, Ad has five ancestors and Bd has also five, but as they have one ancestor in common, the father of the family, they are not separated by 5+ 5 ancestors, but by 5+5 − 1, that is by 9. Similarly, one being always subtracted, the third cousins A and B are related in the seventh ách, and the second cousin in the fifth âch so with the others in odd numbers downwards, and also with the relatives reckoned upwards to the seventh or eighth generation, which would mean collaterals separated by eleven or thirteen ancestors respectively. This reckoning, which is purely conjectural, is based chiefly on the Kenfig story, which foretold the vengeance to come in the ninth âch and otherwise in the time of the goresgynnyd, that is to say in the sixth lifetime. This works out all right if only by the ninth âch we understand the generation or lifetime when the collaterals are separated by nine ancestors, for that is no other than the sixth from the founder of the family. The Welsh version of the ILynclys legend fixes on the same generation, as it says yn oes wyrion, gorwyrion, esgynnyd a goresgynnyd, 'in the lifetime of grandsons, great-grandsons, ascensors, and their children,' for these last's time is the sixth generation. In the case of the Syfadon legend the time of the vengeance is the ninth cenhedlaeth or generation, which must be regarded as probably a careless way of indicating the generation when the collaterals are separated by nine ancestors, that is to say the sixth from the father of the family. It can hardly have the other meaning, as the sinning ancestors are represented as then still living. The case of the Tyno Helig legend is different, as we have the

time announced to the offending ancestor described as amser dy wyrion, dy orwyrion, a dy esgynydion, 'the time of thy grandsons, thy great-grandsons, and thy ascensors,' which would be only the fifth generation with collaterals separated only by seven ancestors, and not nine. But the probability is that goresgynydion has been here accidentally omitted, and that the generation indicated originally was the same as in the others. This, however, will not explain the Bala legend, which fixes the time for the third generation, namely, immediately after the birth of the offending prince's first grandson. If, however, as I am inclined to suppose, the sixth generation with collaterals severed by nine ancestors was the normal term in these stories, it is easy to understand that the story-teller might wish to substitute a generation nearer to the original offender, especially if he was himself to be regarded as surviving to share in the threatened punishment: his living to see the birth of his first grandson postulated no extraordinary longevity.

The question why fairy vengeance is so often represented deferred for a long time can no longer be put off. Here three or four answers suggest themselves :

1. The story of the Curse of Pantannas relates how the offender was not the person punished, but one of his descendants a hundred or more years after his time, while the offender is represented escaping the fairies' vengeance because he entreated them very hard to let him go unpunished. All this seems to me but a sort of protest against the inexorable character of the little people, a protest, moreover, which was probably invented comparatively late.

2. The next answer is the very antithesis of the Pantannas one; for it is, that the fairies delay in order to

involve all the more men and women in the vengeance wreaked by them: I confess that I see no reason to entertain so sinister an idea.

3. A better answer, perhaps, is that the fairies were not always in a position to harm him who offended them. This may well have been the belief as regards any one who had at his command the dreaded potency of magic. Take for instance the Irish story of a king of Erin called Eochaid Airem, who, with the aid of his magician or druid Dalán, defied the fairies, and dug into the heart of their underground station, until, in fact, he got possession of his queen, who had been carried thither by a fairy chief named Mider. Eochaid, assisted by his druid and the powerful Ogams which the latter wrote on rods of yew, was too formidable for the fairies, and their wrath was not executed till the time of Eochaid's unoffending grandson, Conaire Mór, who fell a victim to it, as related in the epic story of Bruden Dáderga, so called from the palace where Conaire was slain 1.

4. Lastly, it may be said that the fairies being supposed deathless, there would be no reason why they should hurry; and even in case the delay meant a century or two, that makes no perceptible approach to the extravagant scale of time common enough in our fairy tales, when, for instance, they make a man who has whiled ages away in fairyland, deem it only so many minutes 2.

1 See the Book of the Dun Cow, fol. 99" & seq.

2 For instances, the reader may turn back to pp. 154 or 191, but there are plenty more in the foregoing chapters; and he may also consult Howells' Cambrian Superstitions, pp. 123-8, 141-2, 146. In one case, p. 123, he gives an instance of the contrary kind of imagination: the shepherd who joined a fairy party on Frenni Fach was convinced, when his senses and his memory returned, that, although he thought he had been absent so many years, he had been only so many minutes.' The story has the ordinary setting; but can it be of popular origin? The Frenni Fach is a part of the