This is first illustrated in a German print of "a fierce mountain Scot" in 1767-one of a pack of cards printed in Nürnberg the Ancient. It shows the tartan trews, and is convincingly " fierce," if not minutely accurate in detail. However, we find in August, 1538, a description of James V. clad in "Highland tartan hose" of three ells, at 48. 4d. an ell. In 1639 the Scotch cavaliers entered England. They are described by Defoe (probably on authentic information) "as like a regiment of MerryAndrews," and Mr. MacRitchie suggests that the harlequin's dress is a survival of the close-fitting trews and coat of a Celtic juggler. In 1723 Macky gives a description of Highland gentlemen at the fair of Crieff "in their slashd short waistcoats, a trousing (which is breeches and stockings of one piece)." The common men "spake all Irish-an unintelligible language to the English." The paper finishes with a note from "Les Celtes," by M. D'Arbois de Jubainville, regarding the use of trousers by the Gauls in the third century before Christ. He asserts that the Gauls derived them from the Germans, they from the Scythians, and the last from the Iranians of Persia. He also derives the Gaelic "triubhas" from old French "trebus," Low Latin "tubrucus," i.e. u-brucus, or high-breeches. Shakespere speaks of the "straight trossers" of the kernes of Ireland (Henry V., Act ш., Sc. 7). We could have wished a little more study as to Irish side-lights, but welcome the paper as a contribution of no little interest on Celtic costume.

* Studies in Early Irish History (from the "Proceedings" of the British Academy, vol. i.). By Professor John Rhys.

PROFESSOR RHYS contributes a valuable Paper to the first volume of the "Proceedings" of the British Academy. It was originally intended as an account of the famous bilingual stone at Killeen Cormac, near Colbinstown; but having found what he takes to be a form of the ancient name of Ireland in the Roman inscription, the Professor proceeds to deal fully with it and other kindred forms of the name, and gives us incidentally an ethnographic sketch of early Ireland, all of which is extremely interesting and useful, whether or not we are satisfied with his reading and interpretation of the stone. Many of the positions have, no doubt, been taken by him before; but here some important modifications and additions are made, and the foundations are more amply secured.

As to the stone, it was Father Shearman who, in 1860, first saw the Roman epigraph. He read it IV VERE DRYVIDES, and translated it "Four true Druids." Unfortunately he at once jumped to the conclusion that he was looking at the tombstone of Dubthach maccu Lugair, and he

saw in the four true druids Dubthach and his three sons. This view he afterwards supported with a great deal of learning, but it seems clear that he was biassed by his pre-conceived idea. He identified the district as that of the Ui Cormaic, probably rightly. He took Killeen Cormac to stand for Cill Fine Cormaic, which seems a needless analysis, except that he later on tried to identify the church supposed to have stood there with the Palladian church called Cill Fine. He found that the Ui Lugair, to whom Dubthach belonged, were a sub-tribe of the Ui Cormaic, and he supposed that Killinulugair, a church (or place) mentioned in certain twelfth-century documents as belonging to Glendalough, was another name for Killeen Cormac. Lastly, he found a reference in the Lebar Brecc to the burial of Dubthach and his three sons in dinlatha ceneoil Lugair, which he translates," at the marshes of Cinel Lugair." (See his "Loca Patriciana," printed in the twelfth and thirteenth volumes of our Journal.) This reference, however, Professor Rhys shows, by comparison with the corresponding passage in the Book of Leinster, to be mistranslated. There the word dinlatha appears as dind flatha. Similar variants in spelling, we may add, will be found in the two Mss. of the Prologue of the "Calendar of Oengus," 1. 155, as printed by Dr. Whitley Stokes.

Professor Rhys accordingly translates the phrase "in the dinn (or fortified height) of the ruling family of the race of Lugar"-a description which, he says, can hardly refer to a church. The correction is, however, not very important, as it is quite doubtful whether there ever was a church at Killeen Cormac. We might just as well suppose the mound to have been originally a fortified place, which, if given up to burial, might have long since lost all trace of fortification. More fatal to Father Shearman's conjecture is the fact that his reading of the Ogam inscription cannot be maintained. He read it Duftano Safei Sahattos, which, with the help of Dr. Stokes, he translated "(Stone of) Duftan, the wise Sage." Unfortunately, there seems to be no doubt that the scores which he took to represent the D and the t of Duftan were no part of the original inscription, and they have consequently been ignored by Brash and later epigraphists. Professor Rhys's reading, in which he has been closely followed by Mr. Macalister, is Ovanos Avi Ivacattos, i.e. (the monument) of Ovanus, descendant of Ivacattus.

The Roman epigraph is, however, the more important. There is some doubt whether it should be read IVVENE DRVVIDES or IVVERE DRVVIDES. Mr. Macalister preferred the former, and ingeniously suggested an equation with the Uvan, as he read it, of the Ogam; but Professor Rhys is almost certainly right in returning to the reading IVVERE, the top of the R having flaked off. The Professor, with some doubt, takes druvides as a genitive singular, and the whole as equivalent to Iverae Druidis; and this he understands to mean "of the Druid of Ireland," and to refer to the man mentioned in the Ogam.

Jour. R.S.A.I. Vol. xxxiv., Consec. Ser. (Vol. XIV., Fifth Series.

2 E

But was Ivvera a name by which Ireland was known? In the first place, Professor Rhys sees in it the native name which gave rise to Insula Sacra, by which, according to Avienus, a writer of the fourth century, Ireland was known to the ancients. M. Gaidoz had already asked, "What more natural than that a Greek writer, thinking he had found the abode of the blessed in this ocean-isle of which only the name was known, should have turned 'Iépvn, or 'Iepvìs vñσos, into 'Iepà vñσos?" Professor Rhys replies, "There was something even more natural than this, namely, that Ivvera should become 'Iépa and then 'Iepá, just as lvverna has corresponding to it in Greek 'Iépvn." This certainly gives a better foundation to the ingenious conjecture of M. Gaidoz. Ireland, it appears, must henceforth be satisfied with being the Insula Sanctorum of early Christian times, and must give up all claims to pre-Christian sanctity.

=

Incited, perhaps rashly, by these brilliant conjectures, the present writer feels inclined to push the matter further. Carnsore Point was known to Ptolemy as 'Iepòv aкpov, the Sacred Promontory, a name always rather puzzling, and now that the Insula Sacra is gone, more puzzling than ever. May we not suppose that it, too, the first point in Ireland which traders from Gaul or South Britain would sight, originally received a name merely meaning "the Irish Head," and that its actual form is due to the same process which transmuted Ivvera into 'Iepá? With reference to this promontory, Camden says: "I don't question but it was called to the same sense (i.e., sacred) by the inhabitants. For the last town in it, where the English landed when they first invaded this island, is called in Irish Banna, which signifies holy."" Camden, no doubt, was told of the word beannuigte, meaning 'blessed,' or 'holy'; but, unfortunately for the argument and the dignity of the place, the Irish name for Bannow Bay (the place intended) is Cuan an bainb the Harbour of the Pig,' a name attested by Keating. (See the first volume of our Journal, p. 191.) This leads to a further conjecture, which we shall offer for Professor Rhys's consideration. Later on in his Paper, in dealing with the triple division of Ireland suggested by the three queens of the Tuatha Dé Danann, namely, Eriu, Fodla, and Banba, he naturally associates Eriu, with the Erna, or Ivernians, of Munster, Fodla with the Ulaid and their ancestor, Ollam Fodla, in ancient Ulster, while the intermediate country is left for Banba. Her name should mean 'the Boar Lady,' or 'Boar Goddess.' This suggests to Professor Rhys that Banba's people may have been an offshoot from the Aestii, a Celtic-speaking people, located somewhere in the fens and islands between the Rhine and the Elbe, and described by Tacitus as "wearing as a religious symbol the device of a wild boar." With the aid of the Welsh story of Kulhwch and Olwen, he is able to locate the Boars at a place called Esgeir Qervel, somewhere in Leinster, whence they crossed to Porth Clais, near St. David's, to fight against King Arthur. "This," he says, "makes for the

association of Banba with Leinster." Now our suggestion is that Bannow, Cuan an Bainb= 'the Boar's Haven,' got its name from these people of the boar-totem. It is about the nearest port in Ireland to the coast of Dyfed, and it is the place where some untold centuries afterwards the men of Dyfed under their Cambro-Norman leaders came to take a more than ample revenge. It is further worth noticing that on the coast, about five or six miles to the east of Bannow, is the parish of Kilturk, anciently called Kenturk (see Journal, vol. iii., p. 218), Cenn Tuipc = 'the Boar's Head'-a name by which we suspect Crossfarnoge Point was designated in pre-Christian times; and that in the same district is the townland of Ballyseskin, baile rescinn = 'the Town of the Marsh '— a name which recalls Sescenn Uairbeóil, with which Esgeir Oervel has been identified by Professor Kuno Meyer. We may be unduly prejudiced in favour of our own little bonneens, but these curious names in the precise place indicated seem a strong confirmation of Professor Rhys's location of Banba and her boars.

To return to IVVERA.-A discussion of this name leads Professor Rhys to a very full analysis of all related names of Ireland, or of Irish peoples, and their assumed forerunners. The general conclusion is reached that the basis of the geographical names was usually some form or other of the ethnic names, as in the well-known case of Scotia, so named from the Scotti. The name Iberi for Goidels is attested in two letters of Columbanus. The Goidelic singular would be Ivera-s, Brythonic Ivero-s, with the genitive and plural Iveri, common at an early date to both. This name is then considered in relation to the eponymous ancestors in the Milesian pedigree, and is found to yield Iar, with a genitive Hir, or Ír, used also as nominative, and then becoming indeclinable. Then lar was provided with a new genitive, Iair, or Ier.

This analysis of the name seems to explain all the forms which at first sight are very confusing; but further, the bifurcation of the name had a curious influence on the pedigrees, and we find Iar or Hír, son of Ith, uncle of Mílid, eponymous ancestor of the Erna of Munster, and Hír, or Ír, son of Mílid, ancestor of the Fir Ulaid, or original inhabitants of Ulster. But Professor Rhys regards both peoples as pre-Celtic, and presumably of the same race, and there are indications in the legend that at one time they were regarded as descendants of one undivided ancestor, and that they (along with their ancestor) were cut in two by the incoming Goidels.

Finally, this ethnic name Ivera-s, plural Iveri, underlying the Iberi of Columbanus, is shown to be the key to the whole group of allied names; not only to the geographical name Ivera, or Ivvera, underlying the 'Iepá of Avienus, and the "Iptv (perhaps "Iepu) of Diodorus, and found on the Killeen Cormac Ogam, but through adjectival forms to the Welsh Merweryd Mare Hibernicum, the Erna of Munster, the kindred names on Ptolemy's map, and, not to mention latinized and other forms, lastly

=

through a derivative to Eriu itself. Why this primitive ethnic name is not better attested is a problem not fully solved. But the suggestion is made that the name was, perhaps, avoided owing to its evident liability to be confused with Iberus, meaning an Iberian of Spain, with which, after all, it may have had a common origin.

We cannot follow Professor Rhys in his geographical distribution of the early inhabitants of Ireland, and its connexion with ethnology and the pedigrees. We might question his location of Ptolemy's Manapia at Arklow, rather than at Wexford; but we have said enough to show the great importance and interest of this Paper, and we have only space for one further remark.

As the scientific naturalist can reconstruct an extinct animal from a tooth, or a single bone, so the trained philologist can arrive at far-reaching conclusions from seemingly very slight data. Thus, from the fact that Ivvera is spelled with a double v (a fact paralleled by other ogamic forms), coupled with the fact that Juvenal and Pomponius Mela both use an intervocalic vv in Iuuerna, their name for Ireland, Professor Rhys incidentally reaches the conclusion that the Goidels practised writing of some kind or other as early, at any rate, as the first century. If the remaining premiss is sound-viz., that the spelling could not have been indicated to a Roman by the pronunciation, whatever exactly it was, but that the sound would have been represented by a single v or (later) b—the conclusion would seem to follow that the form must have come from "an educated Goidel . . . as the spelling which he had been taught, and for which the pronunciation of his language supplied the reason then, or at some previous time." This conclusion has an important bearing upon other problems than that of Juvenal's spelling.