ii. 15, called daughter of Beli, whom one can only have regarded as

her father. So for the present I continue to accept Stokes' rendering of Tuatha Dé Danann as 'the Folks of the Goddess Danu.'

[241] See the Oxford Mabinogion, p. 102; Guest's trans., ii. 252. The combination occurs also in the Book of Aneurin: see Stephens' Gododin (London, 1888), p. 322.

[242] It will be noticed that there is a discrepancy between the gutturals of these two words: tyngu, 'to swear' (O. Ir. tongu, 'I swear'), has ng--the Kulhwch spelling, tynghaf, should probably be tyngaf--while tynghed and its Irish equivalent imply an nc. I do not know how to explain this, though I cannot doubt the fact of the words being treated as cognate. A somewhat similar difference, however, occurs in Welsh dwyn, 'to bear, carry, steal,' and dwg, 'carries, bears': see the Revue Celtique, vi. 18-9.

[243] See the Oxford Mabinogion, p. 100, and Guest's trans., ii. 249, where it is rendered 'a wife as a helpmate,' which is more commonplace than suggestive.

[244] La Cité antique (Paris, 1864), p. 50; see also Joachim Marquardt's Privatleben der Römer (Leipsic, 1886), pp. 49-51, and among the references there given may be mentioned Dionysius of Halicarnassus, ii. 25.

[245] See Vigfusson and Powell's Corpus Poeticum Boreale, i. 126, 181-3, 197; the Prose Edda in Edda Snorronis Sturlæi (Copenhagen, 1848), i. 90-2, 102, 104, 172-86; and Simrock's Edda (Stuttgart, 1855), pp. 292-3, 295-6, 299, 316-20.

[246] Two versions of a story to account for the Ultonian couvade have been published with a translation into German, by Prof. Windisch, in the Berichte der k. sächs. Gesellschaft der Wissenschaften (phil.-hist. Classe) for 1884, pp. 338 et seq. Sundry references to the couvade will also be found in my Hibbert Lectures, where certain mythological suggestions made with reference to it require to be reconsidered. But when touching on this point it occurred to me that the wholesale couvade of the Ultonian braves, at one and the same time of the year, implied that the birth of Ultonian children, or at any rate those of them that were to be reared, took place (in some period or other of the history of their race) at a particular season of the year, namely, about the beginning of the winter, that is when food would be most abundant. I have since been confirmed in this view by perusing Westermarck's work on the History of Human Marriage, and by reading especially his second chapter entitled 'A Human Pairing Season in Primitive Times.' For there I find a considerable body of instances in point, together with a summary treatment of the whole question. But in the case of promiscuity, such as originally prevailed doubtless at the Ultonian Court, the question what men were to go into couvade could only be settled by the confinement of them all, wherein we have an alternative if not an additional reason for a simultaneous couvade.

[247] See Strabo, iii. 165, and Diodorus, v. 14.

[248] For some more detailed remarks on the reckoning of descent by birth, see The Welsh People, pp. 36 et seq.

[249] In Welsh eli means 'ointment,' probably so called from spells pronounced over it when used as a remedy. In the Twrch Trwyth story (Oxford Mabinogion, p. 138) one of Arthur's men bears the curious designation of Reid6n uab Eli Atuer, which might be Englished 'R. son of the Restoring Ointment,' unless one should rather say 'of the Restoring Enchantment.'

[250] See the Book of the Dun Cow, fo. 128b, and Windisch's Irische Texte, pp. 138-9. The rebirth of Lug as Cúchulainn has been touched upon in my Hibbert Lectures, p. 431; but since then the whole question of rebirth has been discussed at length in Nutt and Meyer's volumes entitled The Voyage of Bran (London, 1895).

[251] Tylor's Primitive Culture, ii. 4, where he gives a reference to Gustav Klemm's Culturgeschichte, iii. 77, and Klemm's authority proves to be Jessen, whose notes are given in a 'tractatus' bound with Knud Leem De Lapponibus Finmarchiæ (Copenhagen, 1767): Jessen's words in point read as follows, p. 33:--Et baptismum quidem, quem ipsi Laugo, i. e. lavacrum appellabant, quod attinet, observandum occurrit, foeminam Lapponicam, jam partui vicinam, atque in eo statu Sarakkæ impensius commendatam, de nomine, nascituro infanti imponendo, per insomnia plerumque a Jabmekio quodam admonitam fuisse et simul de Jabmekio illo, qui, ut ipsi quidem loqui amarunt, in hoc puero resuscitandus foret, edoctam. Hujusmodi per insomnia factas admonitiones niëgost nuncuparunt Lappones. Si gravida mulier a Jabmekio hac ratione edocta non fuerit, recens nati infantis vel parenti vel cognatis incubuit, per to Myran, in tympano, securi vel balteo susceptum, vel etiam Noaaidum consulendo, explorare, quo potissimum nomine infans appellandus esset. In the body of Leem's work, p. 497, one reads, that if the child sickens or cries after baptism, this is taken to prove that the right ancestor has not been found; but as he must be discovered and his name imposed on the child, resort is had to a fresh baptism to correct the effects of the previous one.

[252] See Holder's Alt-celtischer Sprachschatz, s. v. Lugus; also the index to my Hibbert Lectures, s. v. Lleu, Lug, Lugoves.

[253] For more on this subject see the chapter on the Pictish question in The Welsh People, pp. 36-74.

[254] It is right to say that the story represents the fairies as living under the rule of a rí, a title usually rendered by 'king'; but rí (genitive rig) was probably at one time applicable to either sex, just as we find Gaulish names like Biturix and Visurix borne by women. The wonder, however, is that such a line as that just quoted has not been edited out of the verses long ago, just as one misses any equivalent for it in Joyce's English expansion of the story in his Old Celtic Romances, pp. 106-11. Compare, however, the Land of the Women in the Voyage of Maildun (Joyce, pp. 152-6), and in Meyer and Nutt's Voyage of Bran, i. 30-3.

[255] This conclusion has been given in a note at the foot of p. 37 of The Welsh People; but for a variety of instances to illustrate it see Hartland's chapters on Supernatural Birth in his Legend of Perseus.

[256] See Frazer's article on 'The Origin of Totemism' in the Fortnightly Review for April, 1899, p. 649. The passage to which it refers will be found at p. 265 of Spencer and Gillen's volume, where one reads as follows:--'Added to this we have amongst the Arunta, Luritcha, and Ilpirra tribes, and probably also amongst others such as the Warramunga, the idea firmly held that the child is not the direct result of intercourse, that it may come without this, which merely, as it were, prepares the mother for the reception and birth also of an already-formed spirit child who inhabits one of the local totem centres. Time after time we have questioned them on this point, and always received the reply that the child was not the direct result of intercourse.' It is curious to note how readily the Australian notion here presented would develop into that of the Lapps, as given at p. 658 from Jessen's notes.

[257] This feature of Welsh has escaped M. de Charencey, in his instructive letter on 'Numération basque et celtique,' in No. 48 of the Bulletin de la Soc. de Linguistique de Paris, pp. cxv-cxix. In passing, I may be allowed to mention a numerical curiosity which occurs in Old Irish: it has probably an important historical significance. I refer to the word for 'seven men' occurring sometimes as morfeser, which means, as it were, a magnus seviratus or 'big sixer.'

[258] The non-Welsh names of the fairy ancestress ought possibly to lead one to discover the origin of that settlement; and a careful study perhaps of the language of the Belsiaid or Bellisians, if their Welsh has any dialectic peculiarities, might throw further light on their past.

[259] Our stories frequently delight in giving the fairy women fine dresses and long trains; but I would rely more on the Ystrad Meurig smith's account (p. 245), and the case of the Pennant fairy who tears to shreds the gown offered her (p. 109).

[260] The difference between Mod. Welsh cor and Breton korr is one of spelling, for the reformed orthography of Welsh words only doubles the r where it is dwelt on in the accented syllable of a longer word: in other terms, when that syllable closes with the consonant and the next syllable begins with it. Thus cor has, as its derivatives, cór-rach, 'a dwarf,' plural co-ráchod, cór-ryn, 'a male dwarf,' plural co-rýnnod. Some of these enter into place-names, such as Cwm Corryn near Llanaelhaearn (p. 217) and Cwm Corryn draining into the Vale of Neath; so possibly with Corwen for Cor-waen, in the sense of 'the Fairies' Meadow.' Cor and corryn are also used for the spider, as in gwe'r cor or gwe'r corryn, 'a spider's web,' the spider being so called on account of its spinning, an occupation in which the fairies are represented likewise frequently engaged; not to mention that gossamer (gwawn) is also sometimes regarded as a product of the fairy loom (p. 103). The derivation of cor is not satisfactorily cleared up: it has been conjectured to be related to a Med. Irish word cert, 'small, little,' and Latin curtus, 'shortened or mutilated.' To me this means that the origin of the word still remains to be discovered.

[261] For Edern's dwarf see Foerster's Erec, lines 146-274 and passim, the Oxford Mabinogion, pp. 248-61, and Guest's trans., ii. 73-92; and for Peredur's the latter books, pp. 197-9 and i. 304-7 respectively.

[262] The story of Canrig (or Cantrig) Bwt is current at Llanberis, but I do not recollect seeing it in print: I had it years ago from my father-in-law. The statement as to Carchar Cynric Rwth comes from William Williams' Observations on the Snowdon Mountains (London, 1802). The Bwlch y Rhiw Felen legend was read by me to the British Archæological Association at its meeting at Llangollen, and it was printed in its Journal for December, 1878. It is right to say that the Llangollen story calls the woman a giantess, but I attach no importance to that, as the picture is blurred and treated in part allegorically. Lastly, the use of the word carchar, 'prison,' in the term Carchar Cynric Rwth recalls Carchar Oeth ac Anoeth, or 'the Prison of Oeth and Anoeth,' p. 619 above: the word would appear to have been selected because in both cases the structure was underground.

[263] See the Acta Sanctorum, April 11, where one finds published the Latin life written by Felix not long after Guthlac's death. See also an Anglo-Saxon version, which has been edited with a translation by Ch. W. Goodwin (London, 1848).

[264] In connexion with them Mr. Bullock Hall reminds me of Icklingham, in West Suffolk; and there seem to be several Ickletons, and an Ickleford, most or all of them, I am told, on the Icknield Way. The name Icel, whose genitive Icles is the form in the original life, has probably been inferred from the longer word Iclingas, and inserted in due course in the Mercian pedigree, where it occupies the sixth place in descent from Woden.

[265] Since the above was written, Dr. Ripley's important work on the Races of Europe (London, 1900) has reached me, but too late to study. I notice, however, that he speaks of an island of ancient population to the north of London and extending over most of the counties of Hertford, Buckingham, Bedford, Rutland, and Northampton, as far as those of Cambridge and Lincoln. A considerable portion of this area must have been within the boundaries of Coritanian territory, and it is now characterized, according to him, by nigrescence, short stature, and rarity of suicide, such as remind him of Wales and Cornwall: see his maps and pp. 322, 328, 521.

[266] See Fiacc's Hymn in Stokes' Goidelica, p. 127, l. 41.

[267] The Welsh passages unfortunately fail to show whether it was pronounced sidi or sidi: should it prove the latter, I should regard it as the Irish word borrowed.

[268] Skene's Four Ancient Books of Wales, ii. 153-5, 181-2.

[269] For more about Picts and Pechts see some most instructive papers recently published by Mr. David MacRitchie, such as 'Memories of the Picts' in the Scottish Antiquary, last January, 'Underground Dwellings' in Scottish Notes and Queries, last March, and 'Fairy Mounds' in the Antiquary, last February and March.

[270] See p. 424 above, where, however, the object of the Ogams written on four twigs of yew has been misconceived. I think now that they formed simply so many letters of inquiry addressed by Dalán to other druids in different parts of Ireland. We seem to have here a ray of light on the early history of Ogam writing.

[271] See the Book of the Dun Cow, fo. 130b.

[272] See the Book of Leinster, fo. 117a.

[273] Corrguinigh occurs in the story of 'The Second Battle of Moytura,' where Stokes has rendered it 'sorcerers' in the Revue Celtique, xii. 77; and corrguinacht heads an article in O'Davoren's Glossary, published in Stokes' Three Irish Glossaries, p. 63, where it is defined as beth for leth cois 7 for leth laimh 7 for leth suil ag denam na glaime dicinn, 'to be on one foot and with one hand and one eye doing the glám dicenn.' The glám dicenn was seemingly the special elaboration of the art of making pied de nez, which we have tragically illustrated in the case of Caier.

[274] In Appendix B to The Welsh People, pp. 617-41.

[275] See Rosellini's Monumenti dell' Egitto (Pisa, 1832), vol. i. plates clvi, clx, and Maspero's Histoire Ancienne (Paris, 1897), ii. 430.

[276] One may now consult Nicholson's paper on 'The Language of the Continental Picts': see Meyer and Stern's Zeitschrift, iii. 326-8, 331-2, and note especially his reference to Herodian, iii. 14, § 8. For Chortonicum see Die althochdeutschen Glossen (edited by Steinmeyer and Sievers), iii. 610; also my paper on 'The Celts and the other Aryans of the P and Q Groups' read before the Philological Society, February 20, 1891, p. 11.

[277] I am chiefly indebted to my friend Professor A. C. Haddon for references to information as to the dwarf races of prehistoric times. I find also that he, among others, has anticipated me in my theory as to the origins of the fairies: witness the following extract from the syllabus of a lecture delivered by him at Cardiff in 1894 on Fairy Tales:--'What are the fairies?--Legendary origin of the fairies. It is evident from fairy literature that there is a mixture of the possible and the impossible, of fact and fancy. Part of fairydom refers to (1) spirits that never were embodied: other fairies are (2) spirits of environment, nature or local spirits, and household or domestic spirits; (3) spirits of the organic world, spirits of plants, and spirits of animals; (4) spirits of men or ghosts; and (5) witches and wizards, or men possessed with other spirits. All these and possibly other elements enter into the fanciful aspect of fairyland, but there is a large residuum of real occurrences; these point to a clash of races, and we may regard many of these fairy sagas as stories told by men of the Iron Age of events which happened to men of the Bronze Age in their conflicts with men of the Neolithic Age, and possibly these, too, handed on traditions of the Palæolithic Age.'

[278] See the Berlin Zeitschrift für Ethnologie for 1894, vol. xxvi. pp. 189-254, which are devoted to an elaborate paper by Dr. Jul. Kollmann, entitled 'Das Schweitzersbild bei Schaffhausen und Pygmäen in Europa.' It closes with a long list of books and articles to be consulted on the subject.