Open main menu

Lecture I.



The inhabitants of ancient Gaul were the earliest Celts of whose religion we possess any knowledge: the sources of our information are twofold, namely, the testimony of ancient authors and that of votive tablets or other epigraphic monuments. Of the ancients who touch on Gaulish religion, Caesar, in his account of the Gallic War, may be regarded as far the most important for our propose, partly because he wrote at a time when the process of assimilating the gods of Gaul to those of Italy was only beginning, and partly because he, who was pontiff at home, had opportunities of understanding likewise much about Gaulish religion, not the least of which consisted in his having the druid Diviciacus as his constant companion and intimate friend throughout the war; still there are many reasons for accepting Caesar's account of the Gaulish pantheon with great caution. His words, so far as they bear on the individuality and respective rank of what he considered to be the chief divinities of the Gauls, are to the following effect:[1] They worship Mercury, he says, above all others, and of him they have very many images. Their traditions make him the inventor of all the arts and the patron of roads and journeys, and they think him the most powerful in the matter of acquiring money and in the transactions of commerce. After him, they worship Apollo, Mars, Jupiter and Minerva: of these they entertain much the same opinion as other nations, namely, that Apollo drives away diseases, that Minerva teaches the elements of the various trades and arts, that Jupiter rules over the sky, and that Mars has the direction of wars. Indeed, it is usual for them, as soon as they have resolved to engage in battle, to vow beforehand to Mars all the spoils they may take in the war; so they sacrifice to him all the animals captured, and bring all the rest of the booty together to one spot. In many of their cities, heaps of these things may be seen piled up in their sacred places. Nor does it often happen that anybody so far disregards the traditional custom as to dare either to conceal any of the booty at home or carry away any of the booty set aside: in case such a crime is committed, the offender is tortured and most severely punished.

Such is the purport of Caesar's words, and it will be well to see how far Gaulish epigraphy is found to corroborate or correct them. Unfortunately for the study of Celtic religion and philology, few of the monuments of Gaul supply us with inscriptions in the national tongue; and probably all of them, whether in Gaulish or in Latin, date after the advent of the Roman conqueror and the initiation of his policy of assimilating the gods of vanquished Gaul with those of Borne. This policy took a very definite form under Augustus. He as pontifex maximus united the religions of the Roman world; but the manner in which Africa and the East were treated could not be recommended in the case of Gaul and Spain; so, when he undertook to restore the position of the Lares and Penates, he included among them the Gaulish divinities, who were henceforth styled Augusti. The result in each instance was that the name of the Gaulish god came to be treated more or less as a mere epithet to that of the Roman divinity, with which he began to be regarded as identical: thus the Gaulish Grannos became Apollo Grannus, and Belisama became Minerva Belisama, and so in other cases. Nay, the Roman god not unfrequently seized on the attributes of the native one even to the extent of assuming his Gaulish costume and non-classical appearance, as is amply proved by the images extant in great numbers in France: among others, Mercury, instead of retaining the aspect given him by Italian art, appears often in a form which has been found to recall rather the beauty and artistic perfection of the Greek Apollo. The Roman policy which reduced the Gaulish divinities to Lares Augusti did not stop at that point; for the cult of the Roman gods as such had been introduced, and, as it established itself over the country, it brought with it also that of Mithras, Cybele, and other non-Italian gods and goddesses to whom the Roman pantheon opened its doors. Further, it is found that the worship of the Roman and quasi-Roman divinities was conducted under the superintendence of men of good birth, who bore the title of pontiffs, augurs or flamens; but those in charge of the cult of the Gaulish Lares Augusti were usually freedmen, who bore the designation of Seviri Augusti, and had to discharge their office free of expense to the state. In a word, the Gaulish gods and goddesses were reduced in rank and forced, so to say, to become more or less Roman; but they were not banished or in any way proscribed.[2]

To come to the monuments, I may say that they are to be found in the local museums of France, Switzerland, and portions of neighbouring lands formerly or still occupied by the Celts: they are moreover numerous, and the accounts of them are to be sought up and down the voluminous transactions of some scores of provincial societies, whose publications are not always easy to consult. So I find that, in the absence of a complete corpus of the ancient inscriptions of France, I cannot do better than set out from one district, the monuments of which, as far, at least, as concerns the subject of this lecture, have been laid before the public in a manageable form by competent archæologists. The district I have chosen is that which was occupied in Roman times by the Gaulish state of the Allobroges. It lay mostly on the eastern side of the Rhone, stretching from that river to the Alps, and from the Lake of Geneva to the Isère. To this must be added a certain tract on the other bank of the Rhone as also probably belonging to the Allobroges, and covering at least most of the present department of the Rhone.[3] The metropolis of the Allobroges was the city of Vienna, now called Vienne: their country consisted in part of some of the most fertile land in Gaul, and in part of very mountainous regions. The Allobroges were Celts, though their name means 'those of another march or district:' they were so called doubtless by some of their Celtic neighbours, but the name which they gave themselves is unknown. The peoples on the eastern bank of the Rhone formed a confederation, at the head of which stood the Allobroges, so that they may be said to have had the control of the navigation of that river and of the important traffic carried on by means of it. The Allobrogic confederation formed in its turn a member of the larger one headed by the Arverni.[4] Lastly, my principal authorities for the inscriptions found in the country of the Allobroges are Allmer's collection of the inscriptions of Vienne,[5] and a succinct account of the gods of the Allobroges by the late M. Florian Vallentin, one of the best known archæologists of the south of France.


is the god with whom the monuments lead one to begin, and the first inscription to which I would call your attention was found among some Roman ruins near the village of Beaucroissant in the department of the Isère, and it is said to have read: Mercurio Aug(usto) Artaio Sacr(um) Sex(tus) Geminius Cupitus, ex voto.[6] The place of finding is recorded to have been once called Artay, though the name is unknown there now; but the names Artas and Artay occur near Vienne and Grenoble. This hardly enables one, however, to decide whether the god gave his name to one or more of these places, or the reverse was the case; but one is inclined to the former view by the occurrence of Artio as the name of a goddess in an inscription in the museum at Berne,[7] for one can hardly be wrong in associating with Artio' s name such a Celtic word as the Welsh âr 'plough-land'; whence it would seem by no means improbable, that Mercurius Artaius was the Gallo-Roman title of the god called Mercurius Cultor in an inscription from Würtemberg.[8] This would serve to show that Mercury was associated by the Gauls with agriculture, especially ploughing.

The next inscription to be mentioned was found at Ilières, also in the department of the Isère, and the first portion of it reads : Aug(usto) Sacr(um) Deo Mercurio Victori Magniaco Veilauno.[9] Here the god is styled 'August,' as in the other instance, but the less usual epithet of victor is added, which is to be noticed, as he was no mere Mercury in the Latin sense. Then follow in the inscription two words of Gaulish origin, of which Magniaco would seem to be the name of a place, though it must be admitted to lack the support to be expected from the identification of its modern form as the name of a spot in the neighbourhood. The other, Veilauno, even though it should not prove a misreading of Vellavno, cannot but be regarded as practically identical with it: compare such names as Cassivellaunos, which meant the king or ruler of the hanse or league, and Catuvellauni, of the same import as Caturiges; both peoples being wont, as it would seem, to boast themselves lords of battle or war-kings. It is after the analogy of such compounds that the Gaulish element in the Hières inscription is to be read; that is to say, it makes one compound epithet, Magniaco-vellaunos,[10] meaning, as it may provisionally be rendered, king or ruler of Magniacon or Magniacum, in allusion to some place with which the god's name was associated.

Besides the two foregoing inscriptions in honour of a distinctly Gaulish Mercury, there is monumental evidence that there were temples dedicated to the god at no less than twenty-six different spots[11] in the country of the Allobroges. Some of the twenty-six very possibly belonged to the Greco-Roman Mercury of an imported cult; but the majority may perhaps be assumed to have been those of the native divinity.

So far, then, the monuments agree with the purport of Caesar's words in regard to Mercury; and if we now go beyond the boundaries of the Allobrogic state, we shall find them strongly supported both by the distribution of the inscriptions and the number of the statuettes of the god: the latter prove in some instances of very considerable metallic value—such, for example, as the massive silver Mercury dug up in the gardens of the Luxembourg. M. Gaidoz, in his far too brief account of the religion of the Gauls, speaks of the universality of the worship of Mercury among the Gauls, and calls attention to the number of place-names which bear evidence to it.[12] He mentions the following, but the list might be enlarged: Montmercure, Mercœur, Mercoiray, Mercoire, Mercoiset, Mercuer, Mercurette, Mercurey, Merourie, Mercurot, Mercury. Several such names occur on Allobrogic ground, and the department of the Puy de Dôme, so named from the late Latin word podium, a hill or mountain, contains another podium or puy, known as the Puy de Mercœur; and this last designation, accommodated to the habits of another dialect, yields Montmercure, the name of another place. This completes M. Gaidoz's list,[13] and I would call special attention to the last two as it is noticed that the Gaulish Mercury greatly affected high ground and conspicuous positions. Thus it is supposed that there was a temple dedicated to Mercury on Montmartre: it is known that he had one on Mont du Chat,[14] near the blue lake of the Bourget, in the land of the Allobroges; another on Mont de Sène, in the Côte d'Or; and a third of considerable importance on the Donon, one of the more elevated heights of the Vosges.[15] But far the most celebrated one remains to be mentioned: it stood on the summit of the Puy de Dôme, in Auvergne, and its foundations are said to prove it to have been an extensive and costly building. It was in fact the great temple of the Arverni; and for it was probably destined the colossal Mercury in bronze, stated by Pliny in his Historia Naturalis, xxxiv. 18, to have been made by the Greek artist Zenodorus for the Gaulish state of the Arverni. It stood 120 feet high, and the work took ten years to accomplish.[16] The expense connected with the worship was probably borne by the cities of Gaul in common, and the fame of the temple lasted to the time of Gregory of Tours; for he relates in his Historia Francorum, i. 32, how it was destroyed by Chrocus, king of the Alamanni, which according to the historian happened in the time of Valerian and Gallien.[17] A fragmentary inscription discovered on the spot happens to have been set up by certain negotiatores or men of business, and it serves to show that one of the names under which the god received honour there, was that of Mercurius Arvernus.[18] The focus of his cult has to be sought in Auvergne, but we find from votive tablets that he was also known in Bavaria, in some districts of Rhenish Prussia, and on the banks of the Meuse in the Netherlands.[19] With these must be ranked an inscription at Bittburg, in Rhenish Prussia, to—Deo Mercur(io) Vassocaleti.[20]

But to understand the term Vassocaleti, it would be well to study carefully Gregory's words in the passage already alluded to. He, a native of Auvergne, seems to have been well acquainted with the ruins on the Puy de Dôme, and the following is his account of them: Veniens vero [Chrocus] Arvernos, delubrum illud, quod Gallica lingua Vasso Galate vocant, incendit, diruit atque subvertit. Miro enim opere factum fuit atque firmatum. Cuius paries duplex erat, ab intus enim de minuto lapide, a foris vero quadris sculptis fabricatum fuit. Habuit enim paries ille crassitudinem pedes triginta. Intrinsecus vero marmore ac museo variatum erat. Pavimentum quoque ædes marmore stratum, desuper vero plumbo tectum.[21] Now there seems to be no sufficient reason to sever the Vasso Galate of the manuscripts of Gregory from the Vassocaleti of the Rhenish inscription.[22] One should rather correct the former according to the latter, and then the whole becomes intelligible in the light of Gregory's description of the Gaulish temple. For caleti proves to be the genitive of the adjective which is in modern Welsh caled, 'hard,' in older Welsh calet, Irish calath of the same meaning. The other part of the Gaulish term vasso is to be equated with the Welsh word gwas,[23] 'mansion or palace; Irish foss, 'a staying or rest,' of the same origin as the Greek ἄστυ, 'town or city;' Sanskrit vastu, 'a seat or place,' vas, 'to dwell or remain;' Eng. was, were. So Vasso-calet must have meant the hard mansion or hard palace; perhaps one should rather say the hard temple, since it is believed that the Gaulish noun survived in the old French vas, which meant a chapel, church, temple or cloister. As to the building being called hard, one has only to recall what Gregory has left on record concerning its walls of thirty feet in thickness and the solid nature of the structure generally.

Lastly, I should construe Mercurius Vassocaleti somewhat in a Celtic fashion, as meaning 'Mercury of the Vasso-calet,'[24] or the god who dwelt in that temple. Be that so or not, the Vasso-calet was a very remarkable temple; and what is still more remarkable perhaps is, that the god should have been known by the name of this Arvernian temple of his so far away as Bittburg on the Rhine. But besides the fragmentary inscription already noticed as found on the Puy de Dôme, a complete inscription has been dug up there which supplies us with still another way of designating the god. It is said to read: Num(ini) Aug(usto) et Deo Mercurio Dumiati, Matutinius Victorinus D(ono) D(edit).[25]

Now the name of the mountain, Puy de Dôme, or as it is called by the inhabitants of the district simply le Doum, and the epithet Dumias or Dumiates given to the god whose temple adorned the top of it, cannot well be supposed unconnected, and the question arises as to the nature of the connection between them. Now Dôme or Doum is here probably a Celtic word, and it could in that case hardly be doubted that it should be referred to the same origin as the Irish word duma, 'a tumulus or mound of any kind;'[26] but the Irish duma means an early Celtic form, dumjo-s or dumjo-n, and this stem dumjo we seem to have exactly in Dumiati. The Gaulish word as applied to the mountain may have simply meant the top or summit, in which case the epithet of the god would refer to him as the divinity of the top of the Puy; but other explanations are possible, though I do not think it necessary to detain you with an examination of them.

So much as to the god's epithets; but none of the Allobrogic monuments seem to supply us with any of his Gaulish names, while a curious inscription referring to him comes from Thornbury on the Swale, in Yorkshire, where no name or epithet is given: he is described simply as the discoverer of roads and paths. The words are: Deo qui vias et semitas commentus est.[27] There is, however, no great difficulty in identifying him under a Gaulish name. He was called Ogmios, or at any rate that was one of his principal names, and under that we have a very curious account of his attributes from the pen of Lucian, a chatty Greek, who wrote and travelled in the second century of our era. His words are to the following effect, and though they treat him as Heracles, you will at once see that he was no Heracles in the classic sense of that name: The Celts, he says, call Heracles in the language of their country Ogmios, and they make very strange representations of the god. With them he is an extremely old man, with a bald forehead and his few remaining hairs quite grey; his skin is wrinkled and embrowned by the sun to that degree of swarthiness which is characteristic of men who have grown old in a seafaring life: in fact, you would fancy him rather to be a Charon or Japetus, one of the dwellers in Tartarus, or anybody rather than Heracles. But although he is of this description, he is, nevertheless, attired like Heracles, for he has on him the lion's skin, and he has a club in his right hand; he is duly equipped with a quiver, and his left hand displays a bow stretched out: in these respects he is quite Heracles.[28] It struck me, then, that the Celts took such liberties with the appearance of Heracles in order to insult the gods of the Greeks and avenge themselves on him in their painting, because he once made a raid on their territory, when in search of the herds of Geryon he harassed most of the western peoples. I have not yet, however, mentioned the most whimsical part of the picture, for this old man Heracles draws after him a great number of men bound by their ears, and the bonds are slender cords wrought of gold and amber, like necklaces of the most beautiful make; and although they are dragged on by such weak ties, they never try to run away, though they could easily do it; nor do they at all resist or struggle against them, planting their feet in the ground and throwing their weight back in the direction contrary to that in which they are being led. Quite the reverse: they follow with joyful countenance in a merry mood, and praising him who leads them, pressing on one and all, and slackening their chains in their eagerness to proceed: in fact, they look like men who would be grieved should they be set free. But that which seemed to me the most absurd thing of all I will not hesitate also to tell you: the painter, you see, had nowhere to fix the ends of the cords, since the right hand of the god held the club and his left the bow; so he pierced the tip of his tongue, and represented the people as drawn on from it, and the god turns a smiling countenance towards those whom he is leading. Now I stood a long time looking at these things, and wondered, perplexed and indignant. But a certain Celt standing by, who knew something about our ways, as he showed by speaking good Greek—a man who was quite a philosopher, I take it, in local matters—said to me, Stranger, I will tell you the secret of the painting, for you seem very much troubled about it. We Celts do not consider the power of speech to be Hermes, as you Greeks do, but we represent it by means of Heracles, because he is much stronger than Hermes. Nor should you wonder at his being represented as an old man, for the power of words is wont to show its perfection in the aged; for your poets are no doubt right when they say that the thoughts of young men turn with every wind, and that age has something wiser to tell us than youth. And so it is that honey pours from the tongue of that Nestor of yours, and the Trojan orators speak with a voice of the delicacy of the lily, a voice well covered, so to say, with bloom;[29] a for the bloom of flowers, if my memory does not fail me, has the term lilies applied to it. So if this old man Heracles, the power of speech, draws men after him, tied to his tongue by their ears, you have no reason to wonder, as you must be aware of the close connection between the ears and the tongue. Nrv is there any injury done him by this latter being pierced; for I remember, said he, learning while among you some comic iambics, to the effect that all chattering fellows have the tongue bored at the tip. In a word, we Celts are of opinion that Heracles himself performed everything by the power of words, as he was a wise fellow, and that most of his compulsion was effected by persuasion. His weapons, I take it, are his utterances, which are sharp and well-aimed, swift to pierce the mind; and you too say that words have wings. Thus far the Celt.

According to this account, Ogmios, or the Gaulish Heracles, was the personification of what the Greeks understood by λόγος: he was the god of speech and all that conduced to make speech a powerful agency—eloquence and wisdom, the craft of Hermes, and the varied experience of the travelled old man who had seen many peoples and visited many lands. Now if we wished to discover the equivalent of Ogmjos in the languages of the Celts of the British Islands, we should have to suppose the word submitted to the operation of phonetic processes suggested by other words in their respective vocabularies: thus, according to Old Irish phonology, the, j would go and the word must appear as Ogma, as indeed it does, while in Welsh the changes implied would be rather greater: thus it would first become Ogmijos with j, sounded like English y in the word yes, liable to be modified into đ, or the sound of th in the English word this; moreover, the case-termination must go, and if the word happened to have survived among the Welsh glosses of the 9th century, it would have been found written 'ogmiđ' or 'ogmid.' The next stage would be represented with m softened to v and g to gh, sounded like g in the softest pronunciation of the German word sagen, and soon elided altogether, just as sagen not unfrequently becomes saën in colloquial German, with as little or less trace of the guttural consonant left as in the English equivalent say. The use of the word is first attested in the Black Book of Carmarthen, a Welsh manuscript of the 12th century, and the spelling has since then varied, according to the orthography adopted, from ouit, owit, ouyd, ovyd to ofydd, which is the present orthography, the pronunciation being approximately 'ovüđ,' with its second vowel nearly like a German ü. The exact meaning of the word in the earliest passages where it occurs is not easy to fix; but that of 'one skilled or versed in anything, a teacher or leader,' would suit them all.[30] Later, the duties of an 'ovyđ' were said to be 'to improve and multiply knowledge;' and it is now the name of one of the three kinds of graduates or professors recognized by the Eisteđvod, the other two being bards and druids. Thus if I presented myself as a candidate for a degree without having any claims to be considered a bard or a druid, I should, in case I was not plucked by the presiding druid and his bardic assessors, assume the degree of 'ovyđ,' together with a Welsh proper name. In Welsh the equivalent of the Gaulish word Ogmios has always remained an appellative; but not so in Irish, where Ogma figures as the name of one of the Tuatha Dé Danann, as the gods of the Goidelic pantheon are collectively called in Irish. Nor is this all: he is signalized in Irish mythology as the inventor of writing, that is to say of the Ogam alphabet; for Ogma being much skilled in dialects and in poetry, it was he, we are told,[31] who invented the Ogam to provide signs for secret speech only known to the learned, and designed to be kept from the vulgar and poor of the nation. The motive attributed to Ogma is an invention of a comparatively late age, for there was nothing cryptic about the Ogam alphabet; but the allusion to Ogma's skill in poetry and dialects is important, especially as there was not only a mode of writing called Ogam, but also a kind of pedantic jargon which bore that name.[32] Now Irish legend will have it that the Ogam was so called from the name of Ogma, which is etymologically impossible; so we are left to conclude from the relation in which the words stand to one another, that Ogma was so called from Ogam or that with which he had to do. Supposing that the latter word, the meaning of which is only a matter of inference, signified a letter or a written character, then Ogma would mean he who had to do with writing—the inventor, let us say, of writing; but that is inadmissible, as the Celts probably had no knowledge of writing when the god was first called Ogmios. So we have to look for the key to the meaning of the word Ogam in the direction of spoken rather than of written language. In Scotland, Ogmic writing does not appear to have become known till it was nearly going out of use in Ireland; so one is not surprised to find that in Scotch Gaelic the word Ogam, which is there written oidheam, had no technical meaning, its ordinary significations being that of 'a notion of anything, an idea, inference, meaning, hint;' to which are to be added that of a 'book or pamphlet,' which it is also said to have had.[33] We have probably cognate words in the Greek ὄγμος, 'any straight line, a furrow, a swathe in reaping, a path or orbit;' Sanskrit ajma-s, 'a course, run, expedition;' ajman, which had the meaning of the cognate Latin agmen, as when employed in speaking of waters, of boatmen's oars, and of speech.[34] The various conditions of the problem of fixing the meaning attached to the word Ogam, and the word standing in the same order of priority to Ogmios in Gaulish as Ogam does to Ogma in Irish, seem best satisfied by supposing the common noun to have meant a round or train of words, fluent speech or ready utterance. This harmonizes well with the sketch of Ogmios as the old man Heracles of the Gauls, whose talk and ready wit charmed his hearers. Lucian's picture enables one to portray to oneself the wrinkled, sun-burnt face of the travelled old man, who poured forth the stream of his irresistible eloquence, while his eye flashed with delight and kindly interest. Lucian says that he turned towards his willing captives with a smiling face, and we have the same touch preserved in the Irish legend, when it calls the hero Ogma Grianainech, or Ogma of the Shining Countenance.[35] The combining of the attributes of Heracles and Hermes in one personage, which puzzled the Greek traveller, was no passing whim of the Gauls. The view taken of the god by the Celts was even more comprehensive, for we find him in Ireland wearing not only the character of inventor of the Ogam alphabet, but also that of champion of the Tuatha Dé Danann.


This god is placed before us by Caesar simply in the character of a repeller of diseases—Appollinem morbos depellere—and not in that of the sun-god he was believed by the Greeks to be. Nevertheless, it will be seen as we proceed that some of the Gaulish divinities, equated with him on certain of the monuments of Gaul and other parts of the Celtic world, appear to lay a just claim to be regarded as forms of the sun-god. But to come to the monuments themselves, an altar found at a place near Annecy in Haute Savoie testifies to the worship of a Gaulish Apollo called Virotutes or Virotûs. The inscription is imperfect, and now reads only:[36] Apollini Virotuti T. Rutil(ius) Buricus. We have no further information about this god, and it is unfortunate that the interpretation of his Gaulish name or epithet is a matter of mere conjecture. It seems, however, pretty evident that it is a compound of which the first part viro may be the Gaulish equivalent of the Latin verus, Welsh gwir, Irish fir, 'true,' or else of Latin vir, Welsh gwr, O. Irish fer, 'a man.' The preference, if given to the latter, would suggest that the epithet may have meant man-healing or man-protecting, and thus one might be led to expect in the second element of the name of the god a Gaulish word related in point of origin and meaning to the Latin tutor, 'protector or defender;' but the vocabulary of modern Celts fails to render us any aid in this matter: all that can be said is, that there is no evidence that such a word as we want did not exist in Gaulish.

Beyond the boundaries of the Allobroges, the Gaulish Apollo appears to have been known all over the Celtic world, and he bore several names, of which the most important were Maponos, Grannos and Toutiorix. Three inscriptions[37] in honour of Apollo Maponos have been discovered in the north of England, and in one of them, found near Ainstable, in Cumberland, he is called Deus Maponus, without any allusion to Apollo. Fortunately the name Maponos offers no difficulty: it is the same word as the old Welsh mapon, now mabon, 'boy or male child,' which occurs, for example, in a Welsh poem in the Book of Taliessin, a manuscript of the 13th century: it is there applied to the infant Jesus in a passage describing the coming of the Magi to him at Bethlehem.[38] Thus it seems certain that some of the Celts worshipped an Apollo whom they described as an infant, and this is borne out by a group of inscriptions at the other extremity of the Celtic world of antiquity: I allude to the ancient province of Dacia, and especially Carlsburg and its neighbourhood, in Transylvania, where we find him styled[39] Deus Bonus Puer Posphorus Apollo Pythius, Bonus Puer Posphorus or Bonus Deus Puer Posphorus. Our Maponos is in all probability the Bonus Puer attested by these inscriptions.

We come now to the name Grannos: it occurs in the districts formerly inhabited by Belgic tribes and in the basin of the Rhine. Grannos is probably to be referred to the same origin as the Sanskrit verb ghar, 'to glow, burn, shine;' ghṛṇa, ghṛṇi, 'heat, glow, sunshine,' Lithuanian z'erẽti, 'to glow,' English gleam: in point of form, Grannos would exactly correspond to the Sanskrit word ghṛṇa-s already mentioned, but the former had probably the force of an adjective, conveying much the same meaning as the posphorus, 'light-bringing,' in the Dacian inscriptions. Nor indeed does the correspondence between them end here; for we find that an inscription from the neighbourhood of Horburg, in the Haut-Rhin, calls the god Apollo Grannus Mogounus.[40] But the interpretation of the word Mogounus compels me to trouble you with some more glottological details, which I will put as briefly as possible: in the first place, we clearly have in Mogounus, from which is derived by one or two further steps the well-known place-name Moguntiacum, from the shorter form of which, Moguntia or Mogontia, are derived its modern representatives, French Mayence and German Mainz. The original Gaulish comes doubtless from the same source as the Irish for-mach, 'increase,' tór-mag or tór-mach, 'increase, the act of adding to;' Latin magnus, 'great;' German mögen, macht; English may, might and main. But words of this origin vary widely in point of meaning in the different Aryan languages, and one group of them supplies expression for the idea of a youth who is growing or has just grown to the might and vigour of manhood: sometimes a transition from this meaning takes place to that of a boy or young man as a servant or slave, much as in the case of παιδίον becoming the French and English page, or the Welsh gwas, 'young man,' used mostly now in the sense of servant. The words in point from the stem mag are such as the Gothic magu, 'boy,' mavi (for magvi), 'girl;' the old Irish mug (genitive moga), 'servant or slave;' Welsh meu-dwy, 'a hermit,' literally servus Dei; Cornish maw, 'a lad or servant;' Breton maouez, 'a woman.' Kindred words are also copious in the Aryan languages of the East, but their divergence of meaning is very remarkable: thus Sanskrit, dwelling on another kind of increase of strength or importance, presents us with a vocable magha, meaning 'a gift or reward,' and maghavan, which means 'freely giving, a giver,' said especially of one who rewards priests and minstrels with offerings: the same two words existed also in Zend, but in that language they retained a more ancient meaning, maga being used in the sense of size or magnitude, and magavan in that of a young man who is grown up but not married, a bachelor. This brings me back to Mogounus, since magavan corresponds with it letter for letter excepting only the declension; and this difference is probably due to Mogounus being only known to us as used in a Latin inscription.

Whatever maybe thought of this conjecture, the analogy of the words we have just examined brings us round again to much the same idea which we found underlying the word Maponos, namely, that of a boy or youth; and I have very little doubt in my mind that Apollo Grannus Mogounus expressed very closely the same meaning which we found rendered by the words Puer Posphorus Apollo in the Dacian inscriptions which have already been referred to. As the dispenser of light and warmth, Apollo made himself the repeller of disease, and it is quite in keeping with this that the god is found to have been not infrequently associated with spots celebrated for their mineral or warm springs, such as Aix-la-Chapelle or Aachen, the Roman name of which was Aquæ Granni. Several other places derive their name from him, such as Graux, in the Vosges, where an inscription[41] in his honour was discovered; and as the stream called Eaux Graunnes,[42] which receives the hot waters of Plombières in the Vosges; and as Granheim, near Mengen, in Würtemberg, a spot in or near which another tablet[43] to Grannos was found. Lastly, Dion Cassius tells us, lxxvii. 15, how Grannus was invoked as the equal of Aesculapius and Serapis by Caracalla.

Apollo Grannos as a god of medicinal springs cannot be severed from the Apollo Borvo of an inscription[44] at Bourbonne-les-Bains, in the Haute-Marne, which reads Deo Apollini Borvoni et Damonae, &c. The monuments show the name to have had several forms: Borvo and Bormo are said to be attested in central France, Bormanus in Provence, and Bormanicus in Spain;[45] while the god's associate is in some instances called Bormana. Thus, to return to the land of the Allobroges, one inscription at Aix-les-Bains, in Savoie, has been read: Cn. Eppius(?) Cuticus Bor. u(t) v(overat) s(olvit) l(ibens) m(erito); and another: M. Licin(ius) Ruso Borm. u(t) v(overat) s(olvit) l(ibens) m(erito). In both of these it would be natural to regard Bor and Borm as standing for Bormano, unless the preference were to be given in one or both to the female divinity, in which case the full form would be Bormanae. For it is certain at any rate that in another part of the Allobrogic land this goddess had a temple, namely, at Saint-Vulbaz, formerly called Saint-Bourbaz, near Belley, in the Ain, where an altar reads: Bormanae Aug(ustae) sacr(um) Capri(i) Atratinus. . . . . (et) Sabinian(us) d(e) s(uo) d(ant).[46] The two, Bormanus and Bormana, were worshipped at Aix-en-Diois, in the department of the Drôme; while at Bourbon-Lancy (Saône-et-Loire) the pair bore the names Bormo and Damona, as well as Borvo and Damona, as at Bourbonne-les-Bains.[47] Among other places, the god has left his name to Bourbon-l'Archambault, in the department of the Allier, whence the Bourbons derive theirs. The exact relation between the kindred forms Borvo and Bormo, together with Bormanus and Bormana, is not very clear; but it is Borvo, and not Bormo, that is re-echoed by the French Bourbon, Bourbonne; and it is Borvo that has its reflex in the vocabulary of the Celts of modern times: I allude to the Welsh berw, 'a boiling,' berwi, 'to boil;' Irish berbaim, 'I boil, cook, smelt,' which are of the same origin as the Latin fervĕre and fervēre, 'to boil or to be boiling hot.' It does not appear why the Gaulish word was Borvo rather than Bervo, but there can be no serious doubt as to the close kinship of the words mentioned, or the fact that the god received his name in allusion to the hot springs over the bubbling volume of which he was supposed to preside. Whether he was originally identical with the Gaulish Apollo it is impossible to say, but even in case he was, he comes before us in most of the inscriptions considerably disengaged from the Gaulish Apollo, as may be gathered from his having a distinct associate Bormana or Damona. But, on the other hand, a passage in one of the Panegyrics of Eumenius is supposed to refer to the hot springs of Bourbon-Lancy: the author would seem to treat Apollo as the chief divinity of the place, and he describes him as punishing perjury by means of the boiling streams,[48] though the monuments found referring there to Borvo or Bormo make no allusion to Apollo's own name.

Having said so much of the Gaulish Apollo, it would hardly be fair to pass in silence over the female divinity associated with him. Her name was Sirona, sometimes lisped into Đirona, and a monument now in the museum at Munich gives a bas-relief representation of her and Apollo Grannos.[49] The latter holds a very large lyre in his left hand, and what may have been a plectrum in the other, while on another face of the stone stands Sirona in a long dress: she has the general appearance of one of the class of Gaulish divinities called Mothers or Matrons: in her left hand she has a bunch of fruit, and in her right some ears of corn, which she is holding up. What relation she bore to the god we are nowhere told; but there is nothing to suggest that she was his wife, even if his names Maponus and Mogounus did not tend to render such a supposition inadmissible, which I think they do. She was probably regarded as his mother, and she was certainly capable of being treated independently; for there are monuments in honour of her alone. One of these last is surmounted with her bust in bas-relief, and the face seems to bear the appearance of extreme old age. The sculptor can hardly have considered her the wife of Apollo Maponus, nor need he have represented her so aged even as his mother. He had probably a reason for doing so, and this brings me back to her name. It will be seen that, if we discard the ending common to it with such Gaulish names as Epona, Divona, Matrona and the like, we have remaining only the syllable sir, which one cannot help interpreting in the light of the Irish sír, Welsh hir, both of which mean long; it would thus seem that the name Sirona referred to the goddess as one who was held to be aged and long-lived. This may be corroborated by a related Irish name, Siorna Saoghlach,[50] mentioned in the mythic history of Ireland: the epithet Saoghlach means long-lived, but it was probably added on as an after-thought, for Siorna may have already conveyed the same meaning; at any rate, Siorna may be regarded, according to the ordinary rules of Irish phonology, as representing an early Celtic form, Sîronjos. The person called Siorna is said to have been engaged in the government of Ireland for a century and a half; and his entire lifetime may be reckoned as considerably longer. I venture accordingly to regard Siorna's name as glossed by Saoghlach or long-lived, and to treat the goddess Sirona's name in a similar manner. Thus we seem to have in the Celtic Apollo and Sirona the ever-young sun-god and an old goddess: the pair invite comparison with the young Apollo of the Greeks and his mother Leto; but Greek mythology sheds no decided light on the agedness of the mother as represented by Gaulish remains.

The same remark applies to what I take to be the equivalents in Welsh mythology, of which a word must now be said; for it has already been mentioned that Maponos is in Welsh mabon; but it should be added that it also occurs as a proper name in the Welsh story of Kulhwch and Olwen; to be more correct, one should say that the proper name was Mabon mab Modron or 'Mabon son of Modron.' The latter was the mother's name, and is a word of the same origin as the Latin matrōna, though it would have sounded in early Celtic matrŏna, which, as the name of the river in France now called, by a shortened form of the word, the Marne, was its pronunciation. One cannot help suspecting that in Mabon and Modron—the father's name is never mentioned—we have the exact equivalents of Grannos and Sirona, and one's curiosity is at once roused to inquire what Welsh literature has to say about the former. We are, however, doomed in part to disappointment: the few allusions to Modron are so obscure that they have not yet succeeded in teaching us anything definite as to her attributes; but the story of Kulhwch tells us the following things about her son. He was a great hunter, who had a wonderful hound, and rode on a steed swift as a wave of the sea: when he was three nights old he was stolen from between his mother and the wall, no one knew whither: numberless ages later, it was ascertained by Arthur that he was in a stone prison at Gloucester, uttering heartrending groans and undergoing treatment with which Apollo's bondage in the house of Admetus could not compare in severity: Arthur and his men succeeded in releasing him to engage in the mythic hunt of Twrch Trwyth that could not take place without him:[51] and lastly, he distinguished himself by riding into the waters of the Bristol Channel after Twrch Trwyth and despoiling him of one of his trinkets.[52]

The third epithet of the god which has been mentioned was that of Toutiorix, which occurs in an inscription at Wiesbaden containing the datives Apollini Toutiorigi.[53] That neighbourhood, you will notice, is also celebrated for its waters, and the interest attaching to the word Toutiorix is out of all proportion to its single occurrence. It can only mean king of the people, which as applied to the god reminds one of the role of Apollo in the history of the Hellenic race, that gave him the titles of leader and founder—ἀρχητέτης, κτίστης, οἰκιστής The name Toutiorix, for which one would have expected Toutorix, has its modern representative in the Welsh Tudri, old Welsh Tutri: it is also well known among Teutonic nations from the time of Strabo, who gives it as Δευδόριξ, while Byzantine authors preferred Θευδέριχος or Θεοδέριχος; and Latin writers supply us with Theodoricus, whence the form usual in English books, Theodoric, which comes pretty near the Anglo-Saxon spelling Theódric. The corresponding High German is Dietrich, so well known as that of Dietrich of Bern, where Bern is the German for Verona. Now the great historical Teuton of this name was a remarkable king of the Ostrogoths, and conqueror of Italy in the 5th century: Verona was one of his headquarters. But it is found that with his history so much unhistorical matter has been incorporated, that modern authors usually distinguish between the historical man as Theodoric the Great, and a mythical personage to whom the name Dietrich von Bern is left. Many attempts have been made to disentangle the legends from the historical portions of the story of the Teutonic conqueror;[54] but it has never been satisfactorily shown why such and such mythic stories should have attached themselves to this particular man. The inscription alluded to yields the key: the historical Teuton bore one of the names of the Gaulish Apollo, and the eventual confusion of myth and history was thereby made easy. This is borne out by the general similarity between the mythic statements made about Dietrich and what is known in Celtic literature about Celtic sun-gods. Among other things may be mentioned his riding, like Mabon, into the sea after an enemy, who was only enabled to escape by the intervention of a mermaid, who was his ancestress. As one of Dietrich's solar peculiarities may probably be mentioned his breathing fire whenever he was made angry; and, like more than one of the Celtic sun-heroes, he is made to fight with giants and all manner of wild beasts. One of the localities associated with his story is the well-known Drachenfels above Bonn; nor is it beside the mark to mention that Verona was the name not only of a city in Italy, but also one of the ancient names of Bonn,[55] a town which is, like Wiesbaden, situated in the neighbourhood of the Rhine. It has puzzled historians that Theodoric, the grandest figure in the history of the migration of the Teutonic peoples, should appear in the Nibelungen Lied, not as a great king and conqueror on his own account, but merely as a faithful squire of the terrible Attila, whose empire had in fact crumbled into dust before the birth of Theodoric.[56] But from the mythological point of view, the subordinate position ascribed to Theodoric is quite correct, and it serves to show how profoundly the man's history has been influenced by the legend of the Celtic god.


The next god to be mentioned in the order adopted by Caesar is Mars; and an inscription at Chougny, near Geneva, equates with him a Gaulish god called Caturix. It reads thus: Marti Catur(igi) sacr(um), pro salut(e) et incolumitate D. Val(erii) Am(a)ti, Sex. Cr(is)pin(ius) Nigrinus v(otum) s(olvit) l(ibens) m(erito).[57] This form of the god's name is rendered certain by that of an inscription at Stuttgardt in Würtemberg, in which Marti Caturigi[58] is written in full, and by a third instance, namely, one found in the neighbourhood of Yverdon in Switzerland.[59] The word Caturix is a compound, meaning the king of war or lord of battle, from catu, which is in Welsh cad, and in Irish cath, 'a battle,' and rīx, 'a king,' in Welsh rhi and in Irish , genitive ríg: the cognates of both words are so familiar that I need not enumerate them. The plural Caturiges was the well-known name of a Gaulish people; and, transferred to their town, it is now continued in the abbreviated form of Chorges. The Teutonic name of the same etymology was common as that of a man, and in fact is still so: witness the Anglo-Saxon Heađoric, the modern German Hedrich, and other variations of the same compound.

Another Allobrogic inscription gives the Gaulish Mars another name: an altar found at Culoz, near Belley, in the department of Ain reads: N(umini) Aug(usto), Deo Marti Segomoni Dunati, Cassia Saturnina ex vot(o), v(otum) s(oluit) l(ibens) m(erito).[60] Segomo is known to us by other inscriptions at Arinthod[61] in the Jura, at Contes[62] near Nice, at Lyons,[63] and at Nuits in the Côte d'Or. The god's name is found also in Ireland; for with the word netta (in later Irish nia,[64] genitive niath or niadh, 'a champion or warrior'), it forms the personal name Netta-Segamonas, which may be rendered Propugnatoris Segomonis, '(of) Segomo's Champion.' It was a kind of name very congenial to ancient Irish ideas, and it occurs in three[65] distinct Ogam inscriptions in the county of Waterford; it is also to be found in lists of the early kings of Erinn.[66] The exact signification of the god's name Segomo is not easy to fix: it may have meant the strong one, the holder or upholder, the defender or protector, or else the victorious one that overpowers and conquers: all one can feel certain about is, that the word is derived from the root segh or sagh, 'to hold, restrain, withstand, overpower,' from which such words come as the Greek ἔχω, 'I hold or have,' ἔσχον, ἴσχω, ἐχυρός, and the like, also the Gothic sigis and the German sieg, 'victory.' It is clear, however, that such a name would suit the god, whether viewed more especially as the chief of the gods or as a mighty and victorious warrior.

Let us now return to Segomo's epithet Dunates. Here again uncertainty must prevail, whether the word be derived or not from the name of a place; but no archæologist has, so far as I know, been able to identify any place-name in point: so we are at liberty to interpret the epithet in another way and to refer it to the same origin as the dunum, Gaulish dūnon, of such names as Augustodunum or Autun and Lugdunum or Lyons. This dūn- is of the same etymology as the familiar English word town and the German zaun, 'a hedge or field-fence;' but its long vowel was probably pronounced as it is in modern French; for the Welsh equivalent is dîn, 'a fortress or stronghold,' whence dinas, 'a fortress, town or city.' The Irish is dún, of the same meaning, but of a different declension; but Irish has further a derived verb dúnaim, 'I shut or barricade,' and dunad, 'a camp, an array.' Hence it would seem that Segomo Dunates meant either Segomo the surrounder and defender, or else Segomo as the god who presided over the stronghold, the camp and the army, that is to say, a Gaulish Mars Castrensis. Lastly, two inscriptions at Bouhy,[67] in the department of the Nièvre, are dedicated to Mars Bolvinnus, and one of them to Marti Bolvinno et Duna(ti). This is a considerable distance from the place of finding the Allobrogic inscription; so that if the name is to be regarded as a topical epithet of the god, it must refer to some celebrated temple of his, like that of Mercury on the Puy de Dôme; but as no such temple has been heard of, the probability is strengthened that Dunates is to be interpreted in one of the ways suggested.

Mention has already been made of Segomo Cuntinus[68] connected with Conte; there was also a Mars Vintius, who was worshipped at Vence, near Nice, and who gave the former place its name: this is proved by an inscription found on the spot, mentioning Marti Vintio. Vintius, in Gaulish Vintjos, must have meant 'relating to the wind,' as it is of the same origin as the English word, the Welsh gwynt, Latin ventus; but, more exactly, Vintjos is an adjective from ventos, which was probably the Gaulish word for wind, and from Ventjos was produced by a modulation of the vowel the attested form. It is remarkable that the Welsh gwynt, wind, is the exact equivalent, not of the simpler noun meaning wind, but of the adjective denoting the wind-god. Several reasons might be adduced why the wind should be associated with the war-god; among others, it might be suggested that all violent gales that commit general havoc and destruction might not unnaturally be referred to the agency of the god of war. But the wind is not always destructive, not always adverse; it is sometimes the fair breeze for which the mariner whistles. So it happens that Vintios, associated with fair wind, is found identified with Pollux, a god propitious to sailors. This is attested by an Allobrogic inscription[69] on an altar at Seyssel, in Haute-Savoie, reading: Deo Vintio Polluci, Cn. Terentius, Billonis fil(ius) Terentianus, ex voto. Another, in which Vintius stands alone, was found in the Vigne des Idoles, near the castle of Hauteville[70] in the same department, and reads: Aug(usto) Vin(tio) sacr(um), T. Valerius (. . . . .) Crispinus, sacer Vinti præf(ectus) Pag(i) Dia(. . . . .) ædem d(at). The navigation of the Rhone at the present day begins at Seyssel, and in Roman times the mariners of that river formed a powerful and influential body which had its head-quarters at Lyons: one old inscription describes it as a splendidissimum corpus.[71] It is probable that the god Vintios had many temples and altars in that neighbourhood, and the site of one of them is marked out by the name Vence or Vens, borne by a hill near Seyssel, at whose foot stands now a chapel dedicated to the Virgin, who is in great esteem among the boatmen of the Rhone: their ancestors doubtless worshipped and supplicated Vintios on the same spot. The curious instance we have here of a Gaulish god being, so to say, split up into two Latin ones, throws some light on the treatment which the Gaulish deities experienced under the influence of Rome. For we are under no necessity to suppose with M. Vallentin[72] that the Gaulish mind regarded the Vintios at Nice as a separate and distinct deity from the Seyssel one. The Gaulish wind-god was more probably one, whether the wind he granted at a particular moment happened to be fair or foul. There was a mythological reason for associating the wind with the Celtic war-god, as will be seen later: hence the difficulty in rendering his personality in the terms of Latin theology. So long as it was a question of the wind as a violent or malignant agency, the equation of Vintios with Mars would doubtless fit; but when the wind was favourable to the mariner, then Mars was probably thought out of place, which led to the preference for Pollux.

The names and epithets borne by the Celtic war-god beyond the limits of the Allobrogic land are too numerous to be discussed one by one here, and I will only call your attention to a few of them. Several inscriptions in honour of Mars Cocidius[73] have been found in this country; but the meaning of the word Cocidius is unknown, as well as that of a related form Cocosus,[74] which also occurs. A more transparent epithet is Belatucadrus, given the god on monuments also found here.[75] This is a Celtic compound meaning handsome in the slaughter or mighty to kill. The epithet was doubtless meant as a flattering one, acceptable to the god in his character of warrior and slaughterer of his worshippers' enemies.

The next to be mentioned is Camulos, which I hesitate to call an epithet, as it is not a compound and possibly not an adjective, but a noun, and one of the god's proper names, like Segomo. An inscription recording the building of a temple for Mars Camulus has been met with in the neighbourhood of Düsseldorf,[76] and others are known elsewhere[77] on the Continent, while one is preserved in a museum at Glasgow.[78] It is right to say that most of the Roman inscriptions found in this island may be the outcome of the piety of continental Celts, so that the gods in whose honour they were set up were not necessarily worshipped by the natives of Britain; but even here we have evidence of the popularity of Camulos in the name of the capital of the Trinovantes, which was Camulodunon, or the stronghold of Camulos.[79] The meaning of the god's name is regarded as unknown, but a very safe conjecture may be made on that point; for though there is a scarcity of Celtic words to explain it, there can be little doubt that it is to be equated with the Old Saxon himil and the German word himmel, heaven or sky, which etymologists refer to a stem, hem, Aryan kam, inferred to mean 'curving, vaulting or covering over.' Among other words from this origin have been reckoned the Greek καμάρα,[80] 'anything covered or arched over,' such as a vaulted chamber, a covered barge, or a tester bed; Lat. camera, 'a vault, an arch, a chamber;' camurus, 'crooked, curved;' Zend kamara, 'a vault, a girdle;' kamaredha, 'the skull or the head,' with which is connected the Greek μέλαθρον, 'the ceiling of a room or the main beam that bears it, the roof, a house:' this supposes the Greek noun to stand for κμέλαθρον, and to be identical with an attested κμέλεθρον, explained to mean τάς δοκούς, 'the beams or timber of a house.' As a personal name, Camulos has its etymological equivalent in later Celtic in that of Cumall, king-warrior of Ireland and father of the great Finn, whose doings occupy so much room in Goidelic story. The name is to be compared in the first instance with that of Οὐρανός or Uranus and the Sanskrit Varuṇas; but as that of a Celtic Mars one should undoubtedly regard it as a synonym rather of the Greek Zeus or Italian Jove, both of which names were expressive also of the idea of the sky or the heavens. In the light of this explanation it becomes intelligible how the Celtic Mars, associated with the sky, should have to do with the wind, as proved by his Gaulish title of Vintios; and in answer to the question what a god thus associated with the sky should have to do with war, let it for the present suffice to say that throughout the literatures of Greece and Rome, Zeus or Jove was the supreme arbiter of the fortunes of war. It may be hazardous perhaps to construe in the same sense the words from the Rig-Veda about Dyu or Dyaushpitar as a god of mighty works: I refer to a hymn to his son Indra, who mostly superseded him, and the passage is thus rendered by Prof. Max Müller: 'Dyu, thy parent, was reputed strong, the maker of Indra was mighty in his works; he (who) begat the heavenly Indra, armed with the thunderbolt, who is immoveable, as the earth, from his seat.'[81] But there can be no such doubt with regard to the Teutonic Tiu, whose name (Anglo-Saxon Tiu, gen. Tiues; old H. German Ziu, gen. Ziwes; old Norse Týr, gen. Týs) is etymologically identical with Zeus and Dyu, while all the little that is known of him makes him the war-god of the Teutons, before he was surpassed and superseded by Woden: witness the name of the day which Frenchmen and Welshmen call the day of Mars—English Tuesday, Ger. Dienstag, formerly Ziestag, Old Norse Týsdagr and Týrsdagr. The only difference, then, between Sky as the war-god of the early Teutons and that of the Gauls, was that the latter chose to render Zeus, Jove and Dyu, by another word meaning equally the sky or the heavens, and that word was Camulos. The Gauls stood between the Romans and the Teutons: linguistic affinities connect the Celtic languages closely with the Aryan dialects of ancient Italy; but since I began to write these lectures, I have been repeatedly impressed by the striking similarity between the ancient theologies of Celts and Teutons, and we have here an instance in point. There is, however, further evidence to prove beyond doubt the identity of the Teutonic Tiu with the Celtic war-god under another name than Camulos, but the discussion of it must be postponed. Let it suffice for the present that we have discovered the Jupiter of the Celts, and found that Gaulish theology ascribed to him the discharge of functions which the Romans would have regarded as more properly belonging to Mars.

Such a god as I have alluded to must have once been the greatest of all the Celtic gods, the chief of the Celtic pantheon, a conjecture which is favoured by the natural interpretation of some of the attested epithets of the Celtic Mars. Take, for instance, the dative Rigisamo,[82] which occurs in an inscription found in this country, in the county of Somerset. The word seems to be a superlative, meaning most royal or kingly. A still more remarkable epithet was Albiorix, applied to him in an inscription[83] in a museum at Avignon. The compound should mean king or ruler of Albio, a word which may be identified with the Welsh word 'elfyᵭ,' used by Welsh poets in the sense of the world or the universe: so we may suppose that Albiorix signified king of the world. Lastly, the war-god's associate is called Nemetona on the monuments, as, for instance, on one at Bath.[84] She has been identified by M. Gaidoz[85] with Nemon, the wife, according to Irish tradition, of Nét, the war-god of the ancient Irish. Another tradition, however, gives to the latter as his wife the war-goddess called the Mórrígu, which is important, as her name means the great queen. Why she should have been so called has always appeared a puzzle, but it becomes at once intelligible if we suppose her husband as war-god to have been once the supreme or great god of the Goidels: the rôle assigned her by Irish mythology is, caeteris paribus, not very unlike that of Here or Juno; but it is her name that chiefly concerns us at this point. It is further to be noticed that with the Mórrígu Irish literature is wont to associate another war-fury called the Bodb (or Badb) Catha; nor is it clear that the two names may not have originally referred to one and the same mythological being; but, be that as it may, one finds a Gaulish goddess who bore inferentially much the same name as the Irish Bodb Catha, as proved by an Allobrogic altar discovered in the commune of Mieussy in Haute-Savoie. In its present state it reads:[86] Athubodvae Aug(ustae), Servilia Terentia (votum) s(olvit) l(ibens) m(erito). But as the stone is imperfect on the right side, it is conjectured that the full name was Cathubodvae, which has been supposed to stand for Catubodvae. Although our knowledge of Gaulish does not suffice to enable us to show that Athubodva was an impossible form, still Cathubodva appears to coincide in a manner which can hardly be the result of accident with the Irish Bodb Catha, in which we have the compound name analysed. This last meant the Bodb of war and carnage, to whom Irish literature makes frequent reference. The signification of the word bodva or bodb may readily be guessed from the fact that it corresponds letter for letter to the Anglo-Saxon beadu and the Norse 'böđ,' 'war or battle.' This vocable, both in its Celtic and its Teutonic forms, enters freely into the composition of proper names of men; but it will here suffice to mention the Bodvogenus of another inscription. It would mean a man descended from the goddess Bodva; and much the same must have been the import of the Gaulish Bodvognatus,[87] a name to be detected reduced in Welsh to Bodnod. The equation, if well grounded, of the name of the Gaulish goddess with that of the Irish war-fury, would imply that her cult was widely spread, and that she was a considerable figure in the Celtic pantheon, whether she is to be identified or not with the Mórrígu or great queen.

Lastly, the poet Lucan makes us acquainted with another important designation of the war-god, in his well known lines in the Pharsalia, i. 444, &c.:

'Et quibus immitis placatur sanguine diro
Teutates, horrensque feris altaribus Hesus;
Et Taranis Scythicæ non mitior ara Dianæ.'

The name of the first of Lucan's triad occurs in an inscription found in Hertfordshire, which gives the dative Marti Toutati[88] and one cannot help regarding Toutates as a more usual Gaulish form than the Teutates of the best manuscripts of Lucan. The meaning of the name has never been ascertained, but there is no room for doubt as to the group of words to which it belongs: we have had a closely kindred element in the first part of the compound name Toutiorix, and the principal words in point in the Celtic languages of the present day are the Irish tuath, 'a tribe;' Breton tud, used as a plural in the sense of the English word people; and Welsh tud, which has had its meaning shifted from that of a people to its country. Outside the Celtic area the word appears in Italy as Oscan tauta, touto, and Sabine touta, tôta, 'a community: it was also a Lithuanian and a Teutonic word: as the latter it was well known as the term for the German political body called the Diet, and it yields the adjective Deutsch or Dutch, meaning the vernacular language of the Germans as distinguished from the Latin formerly preferred by scholars and pedants: the Anglo-Saxon form was theod, 'a people,' and a foreigner or alien was eltheod, just as it is 'aỻud' in Welsh: the Gothic was thiuda, Old Norse 'thjóđ,' 'a people.' In all these languages the word was feminine, and we should therefore probably be right in assuming that the Gaulish word for a people or community was touta: a derived form toutjus is attested by one of the few inscriptions extant in the Gaulish tongue. It was found at Vaison, and is

preserved in the museum at Avignon. It is to the following effect: Σεγομαρος Ουιλλονεος τοουτιους Ναμαυσατις ειωρον Βελησαμι σοσιν νεμητον.[89] That is to say, Segomaros (son) of Willo, toutias of Nîmes, made Belisama this grove. It is not certain whether toutius meant merely a citizen or some public official among the people of Nîmes. Perhaps the latter view is preferable, and I would suggest that Toutates in a somewhat similar way meant king. We have a Teutonic parallel of the same etymological[90] origin in the Gothic thiudans, βασιλεύς, Norse 'thjoᵭann,' 'a king,' and A.-Saxon theoden, which also meant a king or lord: both the Norse and the A.-Saxon words are found only in poetry, which is an indication that they are very ancient formations, going back probably far behind the time of Ulfilas, as may be shown by approaching the question from another direction: the word touta and its congeners entered into many proper names, and when the Romans had to write these names they represented the Teutonic dental as they did the Gaulish one, as a simple t: witness Caesar's Teutones, Ammianus Marcellinus' Teutomeres, Eutropius' Teutobodus, and Floras' Teutobochus. Now in Teutones or Teutoni we have the plural as given by Roman authors of the word 'thiudans,' 'thjóᵭann' and 'theoden;' and that a people should have given themselves such a name as Teutones,[91] meaning kings, will surprise no one who has noticed such Celtic names as that of the Remi, which signified princes; those of the Caturiges and Catuvellauni, meaning war-kings or battle-princes; and that of the Bituriges, which actually meant Weltherrscher or lords of the world. This explanation of the origin of the modern term Teutonic is doubtless open to the objection of implying that a natural inclination to brag was not quite confined to the Celt.[92]

Before leaving Lucan's lines about the Gaulish divinities, it is right to quote the following words used by an ancient scholiast in reference to the passage: 'The Gauls believe Hesus to be Mercury, since he is worshipped by merchants; and Taranis, the ruler of wars and the greatest of the celestial gods, him who was accustomed formerly to be appeased with human lives, but now glad of those of animals, to be Jupiter.'[93] The scholiast was utterly wrong in the view he took of Hesus, and not much less so in identifying Taranis with the Roman Jupiter; but it was probably the result of no similar blunder on his part, that he represents the Gauls assigning to the king of their gods the superintendence of war as his special province. The chief god of the Celts before the rise of the Celtic Mercury was their god of war: how, then, was a Roman to express this in terms of Latin theology? To be both intelligible and approximately correct, he must either say that, according to the Gauls, Mars was the chief of the gods, or else that Jupiter was the god of war. The latter was the way in which the scholiast chose to put it, and he is supported by certain statues purporting to be those of a Gaulish Jupiter, which represent him as clad like a Roman warrior in a cuirass and a paludamentum: one such was found at Vaison, while another of colossal dimensions was discovered some ten years ago at Séguret in the department of Vaucluse.[94]

Lecture I.



Mars (continued)

All the facts bearing on the history of the Gaulish war-god conspire to prove that he was once the supreme divinity of the Celtic race; and though it is found convenient to term him briefly the Celtic Zeus or Jupiter, it would be more correct to speak of him in terms of Roman theology as a Mars-Jupiter. But the fact of his occupying only the third position of honour in Caesar's time, is weighty evidence to the great progress in the arts of peace and their ideas of a settled mode of life which the Continental Celts had made since the time of their conquering those portions of Europe which they inhabited when they became subject to Rome. The old god associated with the sky was eclipsed by the younger gods, the Gaulish Mercury and the Gaulish Apollo, just as even before the Wicking period Týr had been cast into the cold shade by the rude glories of Woden, a younger god of a many-sided character. But there were abundant traces in Caesar's time of the past greatness of Toutates, nay as late as that of Lucan in the first century, unless I am mistaken in regarding the fact of his giving Toutates the first place in the lines quoted to you as no mere accident. The most important evidence, however, is to be found in Caesar's words, which I take the liberty of bringing under your notice again: 'With regard to Mars,' he says, 'it is usual for the Gauls, as soon as they have resolved to engage in battle, to vow beforehand to him all the spoils they may take in the war; so they sacrifice to him all the animals captured, and they bring all the rest of the booty together to one spot. In many of their cities, heaps of these things may be seen piled up in their sacred places. Nor does it often happen that anybody so far disregards the traditional custom as to dare either to hide any of the booty at home or to carry any away that has been set aside: in case such a crime is committed, the offender is tortured and most severely punished.' The meaning of these words is quite clear: the god's aid and sympathy, nay his active co-operation, were to be engaged by giving him the spoils which his worshippers took from their enemies, and he who failed to give the god his due was held to bring the divine displeasure on the state, which the criminal thereby rendered liable to defeat and ravage: in other words, he became guilty of the most heinous crime possible against the community.

Plenty of parallels may, doubtless, be found among other ancient nations, but I will only call your attention to the familiar case of the Jahveh of the Hebrews as fully described in the Book of Joshua. We read in the 7th chapter that Joshua, in his distress at finding his men defeated in their attack on a small town called Ai, was addressed by the Lord in the following words: 'Israel hath sinned; yea, they have even transgressed my covenant which I commanded them: yea, they have even taken of the devoted thing; and have also stolen, and dissembled also, and they have even put it among their own stuff. Therefore the children of Israel cannot stand before their enemies, they turn their backs before their enemies, because they are become accursed: I will not be with you any more, except ye destroy the devoted thing from among you.' The narrative then proceeds to relate how the Lord assisted in discovering the thief who had defrauded him of the shekels and fine raiment which were his: the transgressor proved to be Achan, a man of the tribe of Judah. The sequel reads as follows, beginning with Achan's confession: 'When I saw among the spoil a goodly Babylonish mantle, and two hundred shekels of silver, and a wedge of gold of fifty shekels weight, then I coveted them, and took them; and, behold, they are hid in the earth in the midst of my tent, and the silver under it. So Joshua sent messengers, and they ran unto the tent; and, behold, it was hid in his tent, and the silver under it. And they took them from the midst of the tent, and brought them unto Joshua, and unto all the children of Israel; and they laid them down before the Lord. And Joshua, and all Israel with him, took Achan the son of Zerah, and the silver, and the mantle, and the wedge of gold, and his sons, and his daughters, and his oxen, and his asses, and his sheep, and his tent, and all that he had: and they brought them up unto the valley of Achor. And Joshua said, Why hast thou troubled us? the Lord shall trouble thee this day. And all Israel stoned him with stones; and they burned them with fire, and stoned them with stones. And they raised over him a great heap of stones, unto this day; and the Lord turned from the fierceness of his anger. Wherefore the name of that place was called, The Valley of Achor, unto this day.'

Thus far of Achan's high treason: whether the Gauls would have involved all the members of his family in his terrible death, one cannot say; but it is clear that they would have regarded his transgression in exactly the same light as the Hebrews did; and Caesar's words suggest the inference that even in his time, when the war-god had been surpassed in popular esteem by the more genial divinities of trade and health, the former still remained the god of the state in a sense in which no other could well have been. It may help us to understand the scrupulous regard for the rights of the god of war entertained by the Gauls, the Hebrews and other nations of antiquity, if we look for a moment at the traces of this feeling which manifest themselves among the civilized nations of modern times: I need only allude to the singing of solemn Te Deums after victory, or to our praying in this country that our Queen 'may be strengthened to vanquish and overcome all her enemies,' and to our adorning our cathedrals with the tattered flags of the foreigner. That 'the Lord is a man of war' is a sentiment by no means confined to the Song of Moses: it is found to be still a natural one; and I need only remind you of the poet Wordsworth's ode for the English thanksgiving on the morning of the 18th day of January, 1816, and more especially the following lines:

<poem>'The fierce tornado sleeps within Thy courts—

    He hears the word—he flies—
    And navies perish in their ports; For Thou art angry with thine enemies!
    For these, and for our errors,
    And sins that paint their terrors,

We bow our heads before Thee; and we laud And magnify Thy name, Almighty God!

    But Thy most dreaded instrument,
    In working out a pure intent,
    Is man—array 'd for mutual slaughter.
    Yea, Carnage is Thy daughter;

Thou cloth'st the wicked in their dazzling mail, And by Thy just permission they prevail; Thine arm from peril guards the coasts Of them who in Thy law delight: Thy presence turns the scale of doubtful fight, Tremendous God of battles! Lord of hosts!'</poem>

I am quite aware that these utterances have been made the subject of severe criticism; but has any one ever shown that they do not accurately portray the public feeling in this country at the time? For the parochial picture of the Almighty they expose to our view, the poet drew not so much on his own imagination as on that of a war-wearied people, and the paints were mixed by the confident hand of a self-commending Pharisaism. That the ancient Celts and Teutons should have agreed at one time in making their war-god their greatest divinity, or their greatest divinity a war-god, need, then, astonish no one who will bear in mind the ever-present tendency of their descendants to treat in much the same way a God whom they regard as infinitely greater. There are reasons, however, for thinking that the war-like attributes of their war-god never led the ancient Celts wholly to forget the other aspects of his being, though it is not to be denied that, as long as they retained the original habits of the Aryan warrior, the martial qualities of their supreme divinity would be likely to attract undue emphasis; and this state of things among them continued probably a considerable time after a more settled mode of life in a more genial climate had set the Greek mind at liberty freely to develop the many-sided character of the Hellenic god identified with the heavens, who, as the Zeus portrayed by a few masterly touches in the Odyssey, may safely be regarded as the grandest product of heathen theology.


An inscription from Morestel, near La Tour-du-Pin, in the department of the Isère, reads: Iovi Baginati, Corinthus Nigidi Aeliani ex vot(o).[95] Unfortunately the epithet Baginates, which may or may not be topical, is of unknown origin; but compare the Zend bagha, 'god,' and the O. Bulgarian bogŭ, of the same meaning.[96] We are no better off in the case of our next inscription, discovered on a small altar at Vienne: Deo Sucello, Gellia Iucund(a) v(otum) s(olvit) l(ibens) m(erito).[97] The dative is likewise written Sucello[98] on a stone found at Yverdun in Switzerland. The name of the god occurs also on a silver ring found at York and inscribed with the words Deo Sucelo.[99] These inscriptions identify the god with no Roman divinity; but that has at last been compensated for by the discovery of a stone at Mainz reading: I(ovi) o(ptimo) m(aximo) Sucaelo, &c.[100] In spite of the variations in the spelling, the same divinity is probably meant in all these instances, and he is identified by the Mainz monument with Jupiter.

It is needless to add that these data do not enable us to guess in what respect Baginates and the other god were supposed to resemble the Roman Jupiter; nor is it by any means clear how Caesar fixed on the fourth god in his list, the fourth in the order of importance and popularity from the Gaulish point of view, as the one to be placed over against Jupiter. It has sometimes been supposed that the thunderbolt must have been the decisive attribute; but M. Gaidoz[101] reasoning from the monuments combats that view, and rightly points out that Caesar confined himself to the words, Iovem imperium caelestium tenere, which tell us nothing direct about the thunderbolt. M. Gaidoz, who has written at great length on the Gaulish God of the Sun and the Symbolism of the Wheel, regards Jupiter originally as the god of light par excellence, and as having become by an expansion of his attributes the god of the sky or the heavens. He entertains the same idea of the Gaulish god represented with a wheel in his hand, while he regards the thunderbolt as a Roman accessory, the Gaulish symbol for thunder being undoubtedly the hammer, as among the Teutons.[102] His conclusions, then, are that the wheel represents the sun; that the Gaulish god with the wheel, whom he identifies with the fourth god in Caesar's list, was the god of the sun; and that, the Romans having no special god of the sun till after Caesar's time, the latter could not avoid identifying him with Jupiter.[103] This view deserves to be carefully studied, and may be expected to lead to a better understanding of the original nature of the chief god of the early Aryans, but I am inclined to doubt its applicability to Gaulish mythology so late as the time of Caesar. On the other hand, it can scarcely be denied that a Roman must have always been ready to identify with Jupiter any Gaulish god associated with the phenomenon of thunder, however symbolized.

But no one has more accurately estimated the value of such identification than M. Gaidoz: he tells us, for instance, that it would not be made under the influence of scientific comparisons; that it was not writers like Macrobius that saw it done, but Caesar, the soldiers and the Roman colonists in Gaul; that it took place as the result of reports which could do justice only to one of the attributes of the god concerned; that it may have been based even on accidental resemblances; and that, in a word, the Gaulish religion as known to us is a palimpsest, in which the new writing allows isolated words of the older hand to be read, but not much more.[104] Later, in speaking of the whole passage devoted to the Gaulish gods by Caesar, M. Gaidoz urges the same view in words to which I could not do justice without quoting them verbatim:[105] 'En voulant juger la mythologie gauloise d'après ce texte, je me suis dit plus d'une fois que nous étions dans la situation des sultanes d'Égypte avec les leçons de musique de Félicien David. C'était au Caire, en 1834, pendant un voyage que douze saint-simoniens faisaient en Orient. Le vice-roi demanda à Félicien David de donner des leçons de musique à ses femmes; mais pour observer les convenances musulmanes, Félicien David devait donner les leçons aux ennuques qui les auraient transmises et répétées ensuite aux sultanes!'

In spite of Caesar's words, then, I cannot help regarding the Gaulish god whom he equated with Jupiter as far from possessing the importance or rank which that equation would suggest; nor is it improbable, after all, that the phenomenon of thunder was treated as one of the forms of his activity; and at this point something must be said on that subject. The Welsh word for thunder is taran, which enables us to identify several god-names in ancient inscriptions. One of them was Taranucus on a monument from Dalmatia, which reads: Iovi Taranuco, Arria Successa v(otum) s(olvit):[106] another was the related form Taranucnus attested by two inscriptions[107] on the banks of the Rhine, neither of which alludes to Jupiter by that appellation, nor indeed need they be supposed to have meant him. Both names seem to be derived from a simpler one, Taranus, borne by a divinity identified with thunder; and Taranucnus, in Gaulish Taranucnos, is formed like the Gaulish patronymics Oppianicnos, 'son of Oppianos', and Toutissicnos, 'son of Toutissos.' Treated analogously, we have to interpret Taranucnos as meaning the Son of Taranus, or Thunder. A curious inscription found at Vienne, the ancient capital of the Allobroges, identifies Jupiter with his thunder and lightning, since it reads: Iovi Fulguri Fulmini.[108] Still more important is one found at Chester many years ago, and now preserved in the Ashmolean Museum at Oxford: it begins with a dedication, I(ovi) O(ptimo) M(aximo) Tanaro.[109] The Vienne inscription may perhaps be of Celtic origin, but I doubt the Celticity of the other: it should rather be regarded as a monument of the piety of a German in the army at Chester in the year 154, to which it belongs. To begin with, there can be no grave doubt as to the identity, roughly speaking, of Tanaro with the English word thunder, for the Anglo-Saxon thunor, gen. thunres, German donner (for an older donar), and the name of the Norse god Thor, nom. Thórr, gen. Thórs, from a stem thonr-. To have identified the god with his thunder cannot have been greatly at variance with the habits of thought ascribed to the Germans of an earlier time, at any rate if one were to be guided by Caesar's statement, vi. 21, as to their positivism: Deorum numero eos solos ducunt, quos cemunt et quorum aperte opibus juvantur, Solem et Vulcanum et Lunam. In any case, the evidence of the name Thor may be relied on.

Now that Thor and a Gaulish thunderer have been brought together, they cannot be allowed to part company at once. The former is known to have been credited with possessing a celebrated hammer called Mjölnir, with which he performed his feats of might, and the word is probably of the same origin as the Welsh malu, 'to grind;' Latin molo, 'I grind,' molina, 'a mill;' English meal; and related words, with a certain option between r and l, occur in the Latin martulus, 'a hammer;' Old Bulgarian mlatŭ, the same, mlatiti, 'to hammer or beat.' Moreover, as the lightning was the hammer or the bolt of the thunder-god, several of the kindred vocables had that meaning, such as Old Bulgarian mlŭnij Old Prussian mealde, and Welsh 'ment,' singular ' meỻten,' 'a lightning.' Thor's manner of using his mighty hammer was to throw or hurl it; and a similar idea underlies the Welsh word 'ỻuched,' 'ỻucheden,' a lightning, which literally means what is cast or thrown, as it comes from the same etymon as 'ỻuchio,' 'to cast or throw.' Here may be mentioned three remarkable terms for thunderbolts, recorded by Dr. Pughe in his Dictionary under the word 'ỻuched:' they are Ceryg y Lluched, 'the stones of the cast or the lightning;' Ceryg y Cythraul, 'the stones of the devil;' and Ceryg y Gythreulies, 'the stones of the she-devil.' Before the thunderer's weapon developed into a hammer, it must have been a stone, more nearly resembling Thor's dreaded weapon.[110] It was hard, however, for a Roman to avoid falling into error in regard to the Gaulish thunder-divinities. Thus the wheel-god, the Celtic Zeus, was, I take it, the greatest of them; but the hammer-god was also one of the number, and it is he that we appear to have in the fourth Gaulish god in Caesar's enumeration. But the fact of his calling him Jupiter and of the Dalmatian dedication to Jovi Taranuco belong to the misleading identifications so felicitously estimated at their proper value by M. Gaidoz. Had we more information about the Gaulish hammer-god, we should probably find him resembling Thor still more strikingly. We are told[111] of the latter that he was a less complex divinity than Woden, that he had a well-marked and individual character, that he was ever associated with Earth, whose son he was, and whose proudest distinction was to be called the mother of Thor. He figures as the friend of man; he was the husbandman's god, whose wrath and anger were ever directed against the evil powers that injure mortals: his bolt destroyed the foul thick blights that betrayed the presence of the wicked ones, and smote through the huge cloud-masses that seemed to be crushing the earth. Lastly, he was the husband of the golden-haired goddess Sif, in whom one recognizes the corn-field divinity of Ceres.

A good deal of this description of Thor would probably have applied equally well to his Gaulish counterpart, and the name or title I am inclined to identify with the latter is that of the second god in Lucan's triad. To begin with its form and pronunciation, it is to be observed that the poet's verse requires the first syllable to be considered long, while some of the manuscripts read Æsus without the aspirate, and, as there is no reason to suppose the word in its Gaulish form to have had a right to it, Ēsus may be taken as a more correct spelling. This is proved by one of the inscriptions on an altar dug up at Nôtre-Dame in Paris, where we have nouns of the second declension written with o, such as tarvos, 'bull,' and Cernunnos, the name of another Gaulish god, while the one which here concerns us duly appears as Esus of the U declension. The fact strongly corroborates the view of Pictet and others who connect Esus with the Sanskrit asu-s, 'the breath of life, life both as a force and as a condition;' Zend aṅhu-s, 'a lord or master, also world and place generally,' aṅhva, 'one's own self or individual existence, the soul;' Old Norse áss, genitive ásar, plural æsir, 'gods generally, but more especially the older group of Norse divinities,' to which may be added the Anglo-Saxon genitive plural ésa, 'of gods.'

This identification is of great interest, and I venture to mention one or two particulars of a nature to confirm it: the Norse word points to an original nominative ansu-s, the former existence of which is countenanced by Jordanis' allusion to the title of Ansis,[112] which the Goths gave to the deified heroes of their race. On the other hand, Sanskrit and Zend give evidence only to a weakened form, asu-s, and it is with this rather than with ansu-s that Esus seems to go, in so far as concerns its phonology. For the Celtic languages, not unfrequently setting out with the combination es, corresponding to as in Sanskrit, modified it by eliding the sibilant and making the vowel into i, which in Welsh mostly represents a vowel etymologically long: the stages would seem to be ĕs, ez, īz, ī, i, as in the Irish siur or fiur, 'sister;' Welsh 'chwior-, chwioryᵭ:,' 'sisters;' Sanskrit svasār-as, the same; or Welsh tei, now tai, 'houses,' for tegi ( = teges-a) of the same formation as the Greek στέγος or τέγος, 'a roof, a house,' plural (τέγεα) τέγη. According to this conjecture, the Gaulish god's name may have been pronounced Ezus. The word is not known to survive in the modern languages of the Celts as an independent name, but Welsh has a remarkable derivative from it. For just as Sanskrit asu-s yields asura-s, 'living, spiritual, said especially of the gods, of Varuṇa, and of the sky; any spirit or non-material being of an evil nature,' and the Zend aṅhu or ahu yields ahura-s, 'lord or ruler,' as in Ahurô Mazdâo, Old Persian Auramazdâ, the Ormuzd of Milton's great epic, so the Gaulish Esu-s may be regarded as having given rise to a derivative esuro-s, which may be detected in the Welsh iôr, a word meaning lord or ruler, but seldom applied to any one but God. The term has been reduced to a monosyllable pronounced jôr, with a semi-vowel initial indicating, if this equation be well founded, that the first of the three syllables which originally made up the word was not accented by the ancient Celts. A similar remark applies to another title of God in Welsh, namely, Iôn, which is of the same origin and meaning as Iôr, and reminds one of the Old Norse ásynja, 'a goddess.'

We may, then, guess the Gaulish divinity's name to have meant lord, ruler or god; but why should the hammerer have been called lord or god κατ' ἐξοχήν? For the present let it suffice that I cite the analogy of Thor, as he likewise was treated as the lord or Anse[113] par excellence. The terms 'Anse,' 'the Anse of the Country,' and 'the almighty Anse,' always refer, we are told, to Thor in the Old Norse carmina of oaths and vows: the Swedish word åska, 'lightning, thunder,' is explained to mean the careering of the Anse; and the first syllable of such proper names as Norse 'Ásleikr,' Anglo-Saxon 'Óslác,' Norse 'Ásmóᵭr,' Anglo-Saxon 'Ósmóᵭ:,' and Norse 'Ásbiörn,' Anglo-Saxon 'Ósbeorn,' mod. English ' Osborne,' referred to Thor, as esu- probably did to his Celtic counterpart in Gaulish proper names like Esunertus,[114] 'possessed of the might of Esus,' Esugenus, 'offspring of Esus,' and notably Esugenonertus, which must have meant 'endowed with the might of a descendant of Esus,' a term suggestive of a class or group of Esugeni, but whether men or gods must remain undecided, though the singular is found to assume mythological importance in its continuators in Irish and Welsh romance, where we detect it in Eogan and Owein respectively.[115] It is uncertain whether the Gaulish people of the Esuvii or Esubii mentioned by Caesar, ii. 34, iii. 7, v. 24, meant by so calling themselves to claim descent from Esus, since the name may simply be derived from esus as a common noun, meaning a lord or ruler, in which case it would signify the lordly, princely or ruling race, and supply another instance of the brag underlying names of a type already pointed out.

Now where the name Esus occurs, it stands written over a figure of the god, which has been carefully studied by a distinguished French archæologist, M. Robt. Mowat.[116] He describes the bas-relief as representing the god clad in a short tunic, tucked round his waist so as not to impede the free action of the body. He brandishes a square, short-hafted axe, with which he is felling or lopping a tree, the lance-like form of the leaves of which show it to be a willow such as must have grown in abundance on the banks and islands of the Seine. M. Mowat classes this figure with the bronze images and bas-reliefs of the god known by his Latin name as Silvanus. Other representations make him hold in one hand a branch which he has just cut off a tree with a woodman's bill, while a great many monuments give him as his attributes a hammer and a goblet; but in some instances the goblet is absent, while in others the hammer has smaller hammers growing as it were out of it in tree-like fashion: a remarkable specimen[117] of this kind has been discovered at Vienne. The goblet and hammer sometimes accompany dedications to Silvanus by name; but the variations are too numerous to be enumerated. One of the most remarkable is an altar at Lyons, which brings the hammer and the billhook together: it shows the god using a billhook with his right hand and supporting himself with the other on a hammer with a long handle, while the goblet stands at his feet.

The reasoning of M. Mowat leaves one in no doubt that the Gauls identified Esus with the Roman god Silvanus, who presided over woodlands, clearings and gardens, together with the shepherds' interests. But one group of this class of images has been the subject of another attribution, which has the weighty authority of M. Gaidoz and M. Cerquand: they see[118] in the personage with the hammer and goblet the god of thunder, whose name they take Taranis to have been, and one of the best preserved specimens is a bronze image found at Prémeaulx in the Côte-d'Or. It is described[119] as representing a bearded figure holding a cup in his right hand, while the other grasps the handle of a hammer which stands taller than his own person. His dress consists of a short tunic and some kind of closely fitting trousers: his waist is provided with a thick girdle, which one might be tempted to compare with Thor's so-called belt of strength. The Chester dedication to the German thunder-god shows no trace of the hammer, but only a goblet on one side of the inscription and what appears to have been a rose on the other: the monument is unfortunately in a very bad state of preservation. The museums of the Louvre, Saint-Germain, Lyons and Avignon, says M. Mowat, contain more than a score of images of the same type as that of Prémeaulx; moreover, he states that one was discovered at Metz, and that a ring found at Vendeuil-Caply (Oise) has the same image cut on it. These serve to show that the cult of the god in question was not confined to the south of Gaul.

M. Mowat associates all these with Esus as Silvanus, and adds to them a remarkable altar from Ober-Seebach, near Strasbourg, which represents the god supporting himself with his right hand on a hammer with a long handle and holding a cylindrical vase in the other, while to the left at his feet is seen a dog, the habitual companion of Silvanus,[120] and on the right a female figure in a long robe, with her hand on a cornucopia. In the next place he calls attention to an inscription found at Carlsburg in Transylvania, which reads: Silvano Dom. Terrae Matri, Herculi, Sacrum.[121] Both M. Mowat and M. Gaidoz seem to be in the main right, and the solution of the difficulty is to be sought in the character of the Gaulish god, who was in all probability, like Thor, not only the hammerer but also the friend of the farmer, one most laborious part of whose work consisted in cutting down the woods and forests that confined the domain of the ancient plough and hoe: perhaps it would be more correct to regard him as armed with the thunderbolt or hammer as being, and because of his being, the protector of the farmer. The hammer, doubtless, symbolized thunder and lightning; and possibly the drinking-cup, goblet, vase, poculum, urceus, or whatever you may choose to call it, which has been explained variously as Tibullus' scyphus faginus, Columella's alveolus ligneus, and Vergil's sinum lactis or milk-pail, was the symbol of rain and rustic plenty. The presence of the hammer and the dog, or of the hammer and the woodman's billhook, appropriately represent the two sides of the god's character as the hammerer and the farmer's friend, while the same idea is the key to the meaning of the multiple sledge or branching hammer.

In all the images referred to, the god is said to wear a goodnatured face, and seldom to have the hammer in his right hand ready for action: so it may be inferred that he was chiefly invoked as the protector of the farmer and the friend of the woodman. Though he did not habitually brandish his dread hammer, he was still the owner and wielder of the weapon, which he could handle whenever occasion arose: this was possibly uppermost in the mind of the man who, at the end of a successful boar-hunt, dedicated a temple to Silvanus Invictus at Stanhope in the county of Durham; and possibly the application to Silvanus of the adjective Cocidius, usually reserved for the Celtic war-god, was meant to describe a god more closely resembling Thor in his more warlike moods, a Silvanus such as that described by Livy, Hist. Rom. ii. 7, driving the Sabines to flight by the terrifying voice he caused to issue from the forest of Arsia after a contest between the Romans and Tarquin. The designation Silvanus Cocidius[122] occurs in a dedication at Housesteads on the Roman Wall, and it is noteworthy that it was set up by the prefect of a cohort of Tungrians; but the fact does not prove the god to have been Teutonic: witness the Germans who honoured the Celtic god Maponos in an inscription at Ainstable near Armthwaite: those ancient warriors could have taught a lesson in religious toleration to some of the fanatics of the Teutonic race at the present day.

Another instance showing how another Gaulish god was, so to say, split up into two Roman ones, was brought under your attention in the case of Vintios; and it is not impossible that Hercules in the Carlsburg inscription was meant to stand for Camulos, Toutates, or Segomo, the strong god equated with Mars, in which case, Silvanus Dom(esticus), Terra Mater and Hercules, would be virtually Lucan's triad with the order changed. Be that as it may, the analogy of the treatment of Vintios may, I think, be carried further, and I should be inclined on Italian ground to equate Esus not only with Silvanus, but also with the agricultural god Saturn, whose old-world characteristics remind one of Thor as the 'old Anse.' Allusion has been made to the twofold character of Thor as a thunderer and the farmer's friend, and similarly to Esus; but this may possibly be an inexact way of describing him, since it would perhaps be preferable, as already suggested, to regard him as armed with the hammer in consequence and by reason of his being looked to as the farmer's friend and protector, his thunder being his means of vanquishing the evil powers constituting the farmer's foes. This would leave us free to suppose that thunder and lightning originally and naturally belonged to the divinity associated with the sky, the divinity with whom the Gauls continued to connect the wind, and to whom a Latin inscription gives the name of Mars Vintius. But the god here in question was associated probably not so much with the sky as with the earth; and hence it is that some of his attributes are common to him and the divinity of the earth par excellence, the Gaulish Pluto: so much so, in fact, that M. A. de Barthélemy[123] has tried to prove that the god likened by others to Silvanus should be recognized as the Gaulish Dis Pater of whom Caesar speaks. Somewhat the same opinion has since then been advocated by another distinguished French archæologist, M. Flouest.[124] He takes a view which seems to me to be more in harmony with the rapid advance made by Gaulish archæology within recent years in his country, namely, that the identifications suggested by the other writers mentioned are, from the nature of Gaulish theology, in a great measure compatible with one another.

Lucan's Esus is not to be disposed of without noticing his Taranis. Some of the manuscripts read Taranus, and the same form might be inferred from Taranucus and Taranucnus already mentioned; but the existence of Taranus has recently been placed beyond doubt by the discovery in the south of France of a Gaulish inscription[125] in which its dative, Ταρανοου of the U declension, occurs as the name of a divinity. I should, however, hesitate to substitute Taranus for Taranis in Lucan's verse, as I believe both to have been real names of a divinity associated with thunder.[126] One or both were also probably the Gaulish for thunder itself, and careful study of the cognate Celtic words inclines me to regard Taranis, or Taranus, not as a god but as a goddess, which is countenanced by Lucan's verse in that it institutes a comparison with a Diana (p. 44). The corresponding Goidelic form appears in an Ogmic inscription on a remarkable stone at Ballycrovane, near a bend of the long sea-arm called Kenmare River, in the west of the county of Cork. It reads:[127]

M a q i D e c c e d d a s A w i T o r a n i a s.

Monuments commemorative of persons styled Mac Decet have been found not only in Munster, but also in the county of Kildare, in Anglesey and even in Devon. Awi is a genitive like the Latin fili (for filii), and the nominative plural would, as in Latin, be of the same form; further, Awi Toranias or better Awi Toranjas (with j = Eng. y in yes), would, subject to the known laws of Irish phonology, have to become in later times Úi Torna, and the name was borne by a people so-called in the county of Kerry,

in a district where Abbey O'Dorney has perpetuated the ancient designation; while a certain family called O'Cuirres, connected with the barony of Kerrycurrihy in the county of Cork, are also styled Clann Torna, 'Children of Toranis,' in an old poem,[128] and the name seems to have been pretty widely spread in the kingdom of Munster.[129] The genitive Toranjas implies a nominative Toranis, differing only in its o from Lucan's Taranis, which with its a is probably less original than the Irish one. Now the later form which Toranis should take in Irish would be Toirn, and that is also the nominative which should have as its genitive Torna. But just as the Welsh word gwynt, corresponding etymologically to Vintios, the name of the Gaulish god associated with the wind, has lost all reference to the divinity, and become simply a masculine noun meaning wind, so Toirn, the Irish equivalent of the older Toranis, Gaulish Taranis, has ceased to be a proper noun, and come down to modern times in the signification of 'a great noise or thunder;' and it is noteworthy that it is feminine.[130]

Induced by these and the like considerations to regard Taranis as the name of a goddess, I can of course not identify her with Taranucos or Taranucnos. These names I should regard rather as belonging to Esus, and borne by him as the wielder of the hammer or thunderbolt. Taranis would then take her natural place by his side as his associate; and the mistake, which this way of looking at the question would suppose the ancient scholiast to have committed when he made Taranis into a Gaulish Jupiter (p. 47), becomes easily intelligible. But Zeus was not the only wielder of the thunderbolt even in Greek mythology; both Here and Athene could on occasion make use of that dread missile; and even Typho is known to have handled it, though not with signal success. That the ancestors of the Welsh once associated thunder and lightning with a goddess as well as with a god, is rendered fairly certain by the fact that one of the terms for a thunderbolt is Careg y Gythreulies, the Stone of the She-demon. But when we come to the question of the attributes of Taranis, we are embarrassed by a lack of information; the analogy, however, of Thor helps one to form a consistent theory. For, as in his case, it may be supposed that the associate of Esus was either the Earth in the character of his mother, or else, more probably, some personification of the same origin, but conceived more like Thor's wife Sif, the Scandinavian Ceres of the yellow corn-field.

If, then, the idea has anything in its favour, that Esus—and likewise Thor—was provided by ancient imagination with thunder as a means of defending his friend the shepherd and farmer, it would be natural also that his associate should possess the same means of repelling the powers of evil that attacked the crops; and it may here be mentioned, in anticipation of the remarks to be made on Irish mythology later in these lectures, that it represents the corn-field as the chosen battle-ground where the powers favourable to man make war on those other powers that would blight his crops and blast the fruits of his labour. Possibly one would not be far wrong in supposing Artio to have been the companion divinity of Esus, and Taranis one of her names. The goddess Artio has already been noticed as bearing a name kindred with the epithet Artaius of an Allobrogic Mercury (p. 5), and of the same origin as the English Earth, the Teutonic divinity whom Tacitus, in the Germania, cap. xl, calls Mammun Eartham, 'Terram Matrem.' But the name Artio refers especially to ploughing, and the bas-relief accompanying the inscription on the statue at Berne represents the goddess standing robed and holding a patera in her right hand and fruit in the left, while close by stand an oak and an altar loaded with fruits.[131]


Caesar, in his too brief list of the divinities worshipped by the Gauls, gives the last place to Minerva, to whom he states that they ascribed the initiation of the various trades and arts. What Gaulish goddess he had in view it is impossible to say, and the land of the Allobroges seems to yield no inscription identifying any Gaulish divinity with the Roman Minerva. But one found at Saint Bertrand de Cominges, in the Haute Garonne, mentions a temple of Minerva Belisama;[132] and we have the same name in its Gaulish form of the dative case, βελησαμι, in the Vaison inscription (p. 46), which commemorates the making of a grove for the goddess by Segomaros. A trace of the goddess' name is to be detected in the cognomen read Belismius[133] in a Roman inscription at Carleon on the Usk; and Ptolemy gives a river on the west coast of Britain the name Βελισάμα: it was probably the Ribble. Compare the case of the Dee, which the Welsh always regarded as a goddess, in all probability a goddess of death and war.

Were one to be guided by the apparent similarity of the name Belisama to the first element in that of the god Belatucadros, one might be led to suppose that Belisama's chief concern was war, and that she only resembled Minerva as a war-goddess; but it must be admitted that Caesar's words—Minervam operum atque artificiorum initia tradere—afford no ground for supposing that it was any such a martial Minerva he had in view. Further, if we only turn to Irish literature, we there find traces of exactly such a Celtic goddess as he too briefly mentions: an article in the Irish Glossary, called after the name of Cormac, king-bishop of Cashel in the 9th century, tells us that there was a goddess called Brigit, poetess and seeress, worshipped by the poets of ancient Erinn; that she was daughter of the Irish god known as Dagda the Great; and that she had two sisters who were also called Brigit, the one the patroness of the healing art, and the other of smith-work.[134] This means, in other words, that the Goidels formerly worshipped a Minerva called Brigit, who presided over the three chief professions known in Erinn: to her province in fact might be said to belong just what Caesar terms operum atque artificiorum initia. How largely the prestige of the goddess helped to make the fortune of the saint who took her name, St. Bridget or Bride, it would perhaps be difficult to say, and I pass on to the name Brigit, which makes in the genitive Brigte. This implies an early Goidelic nominative Brigentî, and enables us to identify a presumably corresponding goddess in the Brigantia[135] of Latin inscriptions found here, namely in the country of her namesakes the Brigantes.[136] Add to this that a Gaulish inscription found at Volnay, near Beaune, reads: Iccavos Oppianicnos ieuru Brigindoni cantalon.[137] This literally means that Iccavos, son of Oppianos, made for Brigindo something denoted by the accusative cantalon a word of unknown meaning. In Brigindo we have the name of a divinity probably the Gaulish counterpart of Brigit.

If one may trust these conjectures, we have before us traces of a goddess whose cult was practised in Gaul, in Britain and in the sister island, one whose attributes, so far as we know anything about them, favour the conjecture that she was the Celtic divinity mentioned by Cæsar. To the threefold name here ascribed her by way of conjecture, should be added that Brigit was also called Bríg;[138] in fact, this last seems to have been a favourite Irish name for genius personified: thus there was a Bríg Brethach,[139] whose epithet meant judicial, relating to verdicts or the giving of judgment; while a mythic poet and chief judge of Ulster called Sencha had a daughter Bríg, whose business it was to criticize and correct her father's errors: this Egeria closely resembles, it will be seen, one of the Brigits daughters of the Dagda. In brief, the word bríg meant in Irish pre-eminent power or influence, authority or high esteem; while Welsh has reduced the word to bri, 'renown or high estimation.' Among other words related to the names here in question may be mentioned the Welsh word braint, for an earlier breint, still earlier bryeint, which also occurs, and represents, as it is a feminine, an ancient Brythonic form brigantja, identical with that of the goddess Brigantia of the inscriptions. The Welsh braint means prerogative or privilege, which, involving the idea of power not shared in by all, agrees well enough with the meaning here suggested for Brigantia. The name of the Brigantes was doubtless of the same origin, as was also the old Cornish adjective brentyn or bryntyn, 'noble, free,' a word represented in Welsh by brenhin, 'king,' for an older breenhin, which would imply an early form brigantinos.[140] The idea originally expressed by all these words was that of power or greatness of some kind, whence the derivative ones of freedom, nobility, authority and prerogative; and, so far as we can judge, her names of this origin correctly described the goddess in whom the power of initiating and teaching the arts was supposed to reside, the Minerva of the Celts.


Caesar, in his brief list of the gods worshipped by the Gauls, makes no allusion to Dis; but in a subsequent passage he states, vi. 18, that they believed themselves descended from Dis Pater, a doctrine which, according to him, the druids had taught them. For this reason also they measured the lapse of time not by days but by nights, and calculated the dates of their birthdays, together with the beginnings of their months and years, in such a way as to make the night precede the day.

It is remarkable that the territory of the Allobroges is not known to supply a single inscription equating any Gaulish god with Dis, and so far it would seem as though one might construe Caesar's silence into evidence that the Gaulish Dis was not worshipped. That would, however, be an error, and Caesar's treatment of him is perhaps to be ascribed to the Roman view of Dis as a sombre and inexorable deity honoured on the coasts of the Mediterranean with, comparatively few altars: be that as it may, French archæology has succeeded in identifying the Gaulish god, and in showing at the same time that his cult had very firm hold on the Gaulish mind. M. Mowat[141] is again our guide to the interpretation of another of the early Gallo-Roman altars dug up in Paris, namely, one with the figure of the god surmounted by his name, which in its perfect state is known to have read Cernunnos, a kindred form of which occurs on a wax tablet at Pesth in a mention of a funerary college holding its meetings in the temple of a Jupiter Cernenus.[142] In this last one cannot help recognizing a chthonian divinity corresponding to the Jupiter Stygius of the Romans. The form Cernunnos and the Latinized one Cernenus contain the common stem cern-, which may be assumed to be of the same origin as the native words for the Gaulish horn or trumpet, variously given by Greek writers[143] as κάρνον and κάρνυξ: the Welsh and Irish form is corn, of the same etymology and meaning as the Latin cornu, English horn. How this name suits the god, a glance at the Paris altar suffices to explain; for underneath the word Cernunnos is to be seen, bearded and clothed, a central figure whose forehead is adorned with the two horns[144] of a stag, from each of which hangs a torque. The monument is unfortunately in a bad state of preservation; but the head and shoulders are on such a large scale as compared with the other figures on the same block, that the god cannot have been represented as standing or even as sitting on a raised seat: in fact there is no alternative but to suppose, with M. Mowat,[145] that the god was seated cross-legged on the ground, like Buddha. Granted that posture, we are at once led to connect the whole figure with certain well-known sculptures representing a horned, squatting divinity, such as those found at Rheims, at Saintes,[146] the chief town of Charente-Inférieure, and at Vendœuvres-en-Brenne. The last, which is preserved at Châteauroux, in the department of the Indre, represents the god holding a follis or sack in his lap, and on either side there stand two figures of a diminutive Genius, with their feet planted on the coils of a serpent, while each grasps with one hand either horn of the central personage: the other hand of the one Genius holds a torque, and that of his fellow a purse. A contiguous face of the block shows an Apollo Citharoedus sitting in the posture illustrated by a colossal statue of him at Entrain,[147] in the Nièvre. The Rheims monument[148] represents the horned god squatting on a seat between Apollo and Mercury, who have to stand: in the bend of his left arm, which rests on his knee, Cernunnos holds a bag, from which he pours forth a profusion of acorns or beech-mast,[149] which he helps out with his right hand; they drop down between an ox and a stag, figured below in an attitude of attention; while a rat has been carved above the god's head on the tympanum of the pediment, a detail thought to be of significance, seeing that it is an animal dwelling underground.

Lastly, the block from Saintes[150] has the unusual feature of displaying two groups carved back to back on opposite faces of the stone. In both cases the squatting god holds a torque in his right hand and in the other a purse or bag, which is supported on his knee. He appears to be Cernunnos, though the horns are no longer there to prove his identity, as the monument is imperfect. In the principal bas-relief, his associate sits on a seat near him, with a cornucopia resting on her left arm, while a little female divinity stands close to her. The bas-relief on the opposite face of the block shows the god squatting on a base ornamented with two bucrania: to the left, on a base with a single bucranium, stands a naked god, who supports himself on a club; and on the right, on a base devoid of ornament, stands a goddess in a long robe. It is to be noted that the squatting attitude of the god in these instances has been observed also in the case of a little bronze figure discovered at Autun, now in the museum at Saint-Germain, and in that of certain Gaulish coins, on one side of which is seen a squatting figure holding a torque in its right hand.[151]

Such are some of the principal data for our purpose, and from them I infer that, like Dis and Pluto, Cernunnos was the god of the dead or the nether world. As a corollary, we may regard all three as the gods of riches or the lords of the metallic wealth of the world: this rôle of the classical deities is indicated by their names. Thus Pluto, in Greek Πλούτων, is derived from πλοῦτος, 'wealth,' just as Dis, genitive Ditis, is supposed to be a contraction of dives, gen. divitis, 'wealthy;' though the name Cernunnos did not, at least directly, connote wealth of any kind, the attributes of the god, such as the money-bag and the torques, than which no symbols more expressive of wealth were known to the Gauls, amply made up for it; it appears they also conveyed much the same meaning to the minds of the Germans and the Romans.[152]

There are at least two questions which will have occurred to you respecting the Gaulish Dis: why he squatted and why he had horns. M. Mowat would answer the first by showing that the squatting position was familiar to the Gauls, as proved by more than one ancient author;[153] whence it was that they thought proper to give some of their gods, and especially the one whom they regarded as their universal parent, an attitude which was familiar to them and characteristic of their race; so they also represented him wearing the sagum; other gods might take to Roman ways, but he must remain true to old-fashioned Gaul. With regard to the horns, the same archæologist would account for them by suggesting that the Amalthean horn had, in the form of a cornucopia, single or double, everywhere become the emblem of abundance, and further that the Gauls were in the habit of using the cornucopia as an ornament for the head.[154] Neither answer can be said to go far enough, for it seems probable that both the squatting posture and the horns had a mythological signification reaching back beyond the history of the Celts as a distinct branch of the Aryan family, though we may never be able to find out its precise meaning.

Fortunately there remains one source of light on the genesis and history of Cernunnos which no one, so far as I know, has tried, namely, Teutonic and especially Norse mythology. At the very threshold of the latter, one's eyes light at once on an ancient god, Heimdal, the allusions to whom are, so to say, so scanty that Norse students have never been able to draw a complete or consistent picture of him. This god is briefly described by Vigfusson and Powell[155] as follows: 'An ancient god is Heimdal, from whom the Amals spring. There are strange lost myths connected with him; his struggle with Loki for the Brisinga necklace; the fight in which they fought in the shape of seals. He is 'the gods' warder,' dwelling on the gods' path, the Rainbow. There he sits, 'the white god,' 'the wind-listening god,' whose ears are so sharp that he hears the grass grow in the fields and the wool on the sheep's backs, with his Blast-horn, whose trumpet-sound will ring through the nine worlds, for in the later legends he has some of the attributes of the Angel of the Last Trumpet. His teeth are of gold; hence he is 'stud-endowed.' Curious genealogical myths attach themselves to him. He is styled the son of nine mothers; and as Rig's father, or Rig himself, the 'walking or wandering god,' he is the father of men and the sire of kings, and of earls and ceorls and thralls alike. His own name is epithetic, perhaps the World-bow. The meaning of Hallinskiᵭi [another name of his] is obscure.' Such is a summary of the most important passages referring to Heimdal.

The classics picture Pluto holding in his hands the keys of the nether world, from whose bourne no mortal returns, and Heimdal survived to be transformed into St. Peter with the keys: previous to this, his last stage, he was the porter, watchman or warder of the gods, and as such Loki, the enfant terrible of Norse mythology, makes fun of him sitting in the rain; but this view of the northern gods living together and having occasion for a warder at their gates, is a comparatively late one. So it may be inferred that one of Heimdal's earlier occupations was to sit at the entrance to the nether world, which position, besides partly accounting for his reputation as the most stupid of the gods, might legitimately be compared with the squatting of Cernunnos: Pluto also sat, but it is said to have been on a throne of sulphur. Heimdal was a stud-endowed god, with teeth of gold, which clearly involves an allusion to the metallic wealth of the earth so pointedly symbolized in the representations of Cernunnos. They were both in all probability also the guardians of treasure; and it is in this capacity, doubtless, that Heimdal fought with Loki in his attempt to steal the Brisinga necklace, which was the property of the goddess Freija. But you will perhaps ask whether the fact that Heimdal was called the whitest of the Anses does not overthrow all these speculations. On the contrary, if we examine Welsh literature, we find the king of the fairies and the huntsman who fetches to his abode the souls of the deceased, named Gwyn, that is to say White.

It is known that the ruling families among both Celts and Teutons claimed descent from particular gods, a fact seemingly contradictory of Caesar's statement that the Gauls believed themselves the descendants of their native Dis; but, as in the case of Heimdal, who was reckoned by the Norsemen to be the father of all classes of men, kings and thralls alike, the two views were in a manner consistent, the special descent of the heads of particular families from particular deities being not so much contradicted as covered by the general descent from the god of the nether world. The notion that their Pluto was reckoned by the Gauls the fons et origo of all things, the gods included, is countenanced by Caesar's words, which connected with the god the Gaulish habit of reckoning the night before the day; but precedence was also given the night by the ancient Germans, as we are expressly told by Tacitus in his Germania, cap. xi.; and they did so most likely for the same reason as the Celts, who considered night and death to have existed before light and life. This would explain the myth describing how Heimdal was in the beginning of days born of nine giant maidens, nine sisters or nine mothers, in whom we may see a reference to a nonary week: thus Heimdal proves to be the first offspring of time.

His name must have been epithetic; but he had other names, which, together with his blast-horn, remind one of the horned Cernunnos: I allude to Hallinskiᵭi and Heimdali, both of which are said to mean a ram, which suggests that Heimdal was originally represented as a ram. That he was horned is implied likewise by a curious term in Norse poetry for a man's head, namely, 'Heimdal's sword:' Gretti the Strong so speaks[156] of his own head, and it called forth the explanation that Heimdal had some time or other fought in Samson-like fashion with somebody's head as his weapon; but as it is not called a club or hammer, but hjörr, which meant a sword, also a missile weapon, and even a shield, it is highly probable that the original myth represented him as fighting with no head but his own, the horns on which served him for sword, spear and shield all at once. À propos of Heimdal as a ram, the fact should perhaps be mentioned that the Celtic Pluto and his associate frequently have as one of their attributes a serpent or two with a ram's head.[157]

On the Rheims monument Cernunnos has below him, as we have already seen, a stag and an ox, while the Saintes monument has on it several bucrania: both stag and ox were probably animals sacred to the god, and I am inclined to see a reference to him in the bull represented on the western face of the Gallo-Roman altar[158] in Paris. It bears no god's image or name, but the figure of a bull, with three cranes standing on him, one on his head, one on his withers, and a third on his rump, while above are to be read Tarvos· Trigaranvs·,[159] which doubtless mean a 'three-craned bull.' The beast cannot be regarded as exactly representing the god, as he is adorned with a dorsuale,[160] which marks him out as a victim to be sacrificed: the cranes were probably viewed in the same light, but it is right to add that their number was presumably not a matter of accident; for the idea of a triad appears to have played as important a rôle in ancient Gaul[161] as in Ireland and Wales. Now with respect to Jupiter, the bull and the birds occupy on the block exactly the place which they should in case they referred to Cernunnos; and the reason why his victims take up the room where his own figure might be expected, is probably to be sought in some religious scruple or artistic difficulty which prevented the sculptor from portraying this god, who was so unlike the others as regards both form and posture. Possibly the early date of the altar would warrant our supposing that the bold step had not as yet been taken of attempting any kind of image, at least in northern Gaul, of this unwieldy divinity, whom Gaulish theology had hardly succeeded in anthropomorphizing sufficiently to fit him to figure in a group bearing the stamp of Roman respectability.

All the facts at our disposal tend to show that the chthonian deity of Celts and Teutons was held to have the form of a horned beast, such as a stag, bull, goat or ram, and it is now needless to show why one cannot accept the conventional cornucopia as an adequate explanation of that idea. At the same time it would be rash to say that they had no connection with one another, for the usual account of the Cornu Copiae, or horn of plenty, traces it back to the Greek κέρας Ἀμαλθείας or horn of the goat Amaltheia, from which Zeus was nourished, and in which was to be found all that one could desire. Here we have also a horned beast older than Zeus, and the form of the myth does not compel us to assume that the goat was originally regarded as a she-goat: so it is possible that the Amalthean goat and the horned deities are to be referred to a common origin.

This would, however, not be any answer to the question whence the idea of a horned god of the nether world was derived; one might, for example, look for it in a still cruder manner of regarding him not only as the first offspring of time, but also as the first in point of order in space—that is, as the foundation and upholder of the mass of the universe. In that capacity he may have been originally pictured as a huge elk or a gigantic urus sitting quietly under the weight of the world, save when he shook himself and produced earthquakes. Such a piece of cosmogony would not be without a parallel elsewhere, and it calls to mind such a Gaulish proper name as Urogenonertus,[162] which may be interpreted to mean a man 'endowed with or possessed of the might of an Urogen, that is, of a descendant of Urus,' which possibly meant more than the mere beast described by Caesar as living in the great forests of Germany. Having due regard, however, to the god's connection or identity with the earth, that is to say with the solid ground, one should rather suppose the horns, with which the god was endowed, to be the mythical exponents of the hills and mountains which diversify the surface of the globe.

After this digression, let us return to the data provided us by French archæology: the monuments I have described associate with Cernunnos certain other and younger gods who cluster round him like children by the side of their father. Among others we have found Apollo and Mercury in his company, and there would be no difficulty in explaining this grouping. Thus the Gaulish Apollo was especially connected with, and worshipped near, the mineral springs of the country, those perennial sources of health which poured forth their invigorating volume from the deep realm of Cernunnos; and as to Mercury, who was, among other things, the genius of commerce and money-making, much of his stock in trade, so to say, belonged to the same chthonian deity as the owner and dispenser of the metallic wealth of the world. Another of the figures associated with Cernunnos is represented with a club, which, if meant for a Heracles, would remind one of the Greek myth which makes the god of that name assist Atlas in bearing the burden of the superincumbent world; but it must be confessed that the groupings alluded to are open to the suspicion of having been made in some measure under the influence of classical ideas, including possibly that of the Greek mysteries.

There remains the associate of Cernunnos on the Saintes monument, where she is represented sitting in the ordinary way near to the cross-legged god: she has a cornucopia, which implies that she was regarded there as a benignant goddess; but beyond that, one knows nothing about her, not even whether she should be regarded as identical with the goddess standing on the other face of the stone. But with one or both of these goddesses may perhaps be compared a divinity that figures in the Irish and Welsh pedigrees of the gods. Her Irish name was Danu or Donu, genitive Danaun or Donann[163] (also written with nd for nn). She is treated as a goddess par excellence, in Irish dea, of which the genitive had various forms, such as , dée, déi, déa and dae. So the Irish gods, who are reckoned her descendants, are promiscuously called Tuatha Dé Danann, 'the Tribes of (the) Goddess Danu,' Tuath Déa or Déi, 'the Tribe of (the) Goddess,' and Fir Déa, 'the Men of (the) Goddess.' In Welsh her name takes the form Dôn, and the gods descended from her are accordingly called the Children of Dôn. The more important of them will come under our notice as we go on: they are Gwydion son of Dôn, Gilvaethwy son of Dôn, Amaethon son of Dôn, Govynion or, as he is more usually called, Govannon son of Dôn, the smith, whose name is etymologically equivalent to Goibniu, genitive Goibnenn, that of the smith of the Tuatha Dé Danann, and Aranrod or Arianrhod, daughter of Dôn.

Their father is not usually mentioned, but Arianrhod is called daughter of Beli in the Welsh Triads[164] (i. 40 = ii. 5), whence it may be inferred that Beli was reckoned Dôn's husband. He is usually called Beli the Great, son of Mynogan,[165] and his name figures as that of the king of the Brythons in the golden age of their mythic history. It is also doubtless to be identified with that of Bile, father of Mile[166] and of most of the human inhabitants of Ireland as distinguished from gods and demons. But Bile is fabled to have been king of Spain, so that his descendants are described invading Ireland from Spain: what could that mean? Now visits by the heroes of the Brythons to Hades are, as we shall see in a later lecture, sometimes represented as made to Ireland, and the heroes of Ireland setting out for a similar destination are conversely said to come to Britain. But in some instances Spain appears to have been substituted for Hades. Thus a mythic dog forming a terrible Cerberus killed by the sun-hero Cúchulainn is described as brought from Spain; Cúchulainn[167] receives a visit from a warlike maiden who, having fallen in love with him, comes to see him from the Plains of Spain;[168] and Eogan Mór, another form of the solar hero, is enabled by his leman Eadaoin, or Etin, a goddess dwelling in Great Beare Island, to escape from his enemies to Spain, and to return thence in due time to overcome them.[169] Further, a giant of infernal origin, slain on Mont St. Michel in France by king Arthur, is similarly said to have come to settle there from Spain.[170] So the descent of the ancient Irish from Mile, son of Bile, king of Spain, meant nothing more than what Caesar expressed differently when stating that the druids taught the Gauls that they were all descended from Dis Pater.

The names Bile and Beli corroborate this conjecture, as they are doubtless to be interpreted to mean death or some kindred idea, and to be referred to the same origin as the words mentioned (p. 37) in connection with the name of the Celtic Mars Belatucadros. The meaning of the name of Dôn, Beli's consort, was analogous, for Welsh Dôn, Irish Donu or Danu, are to be referred to the same origin as the English word dwindle, North-Eng. dwyne, 'to fall into a swoon,' A. -Saxon dwinan, 'tabescere,' Old Norse dvína, 'to dwindle, pine away,' Sanskrit dhvan, 'to be hidden, to go out or be extinguished,' dhvânta, 'hidden, dark,' and as a neuter noun 'darkness;' possibly also the Greek word θάνατος, 'death.' On the other hand, the Celtic names are not to be severed from, the Welsh word dyn, 'a human being or man in the sense of homo, not of vir, Irish duine of the same meaning, both of which postulate an early form donjos,[171] meaning literally and etymologically a θνητός: to the early Celt, as to the Greek, man was a mortal, as distinguished from the immortal gods and the ancestors who had taken their departure to the Plain of Pleasure in the other world where death was unknown.

A word must now be devoted to the position of the goddess as regards her consort: Cernunnos was the great figure according to Gaulish ideas, and his associate was apparently of smaller consequence in their sight: did the insular Celts reverse the relative position of the Plutonic pair? When the facts are duly weighed it will be found that there is no evidence to that effect. This view is countenanced by the all but complete absence of any statements as to the nature or attributes of the goddess: she looms darkly in the background as the mother of the gods, and any further predicate about her is to be reached only as a matter of inference. The prominence, on the other hand, given to her name in connection with those of her descendants is to be accounted for, it seems to me, as a survival of the custom of describing persons as the children of such and such a woman, without making any reference to the father's name: other instances of the same kind are not numerous in the mythological tales of Wales; they are more frequently met with in those of Ireland;[172] while in Scotland, that is to say, among the Picts, it was the rule that the father was absolutely of no account in the succession.

To return from this digression to Cernunnos, he has, so far as we have gone, been treated as a horned god; but it would not be right to dismiss him without calling attention to another peculiarity of his figure, as shown by some of the representations of him. Thus, a statuette found at Autun, besides giving him one principal face, has on either side of the head a spot, above each of his ears, fashioned into a small face; so that the god was enabled not only to look forward, but also to see on both sides.[173] Some monuments give his three faces the same dimensions,[174] while others provide him with three complete heads;[175] and the Autun figure combined all his most salient attributes, the horns,[176] the three faces, the cross-legged posture, the torque round his neck, and another resting on his lap and separating two ram-headed serpents.[177] Lastly, the triple head was sometimes considered enough, and the artist made no attempt to give the god a body, the lower portion of the block being utilized for other purposes,[178] such as the representation of the god's associates.

Now this strange god, reduced to a wonderful head, which identifies him with Janus, the Roman god of beginnings and endings, has his counterpart in Welsh literature, though attention has, so far as I can remember, never been called to the fact: I allude to Brân, of whom the Mabinogi speaks,[179] which is called after the name of his sister Branwen, daughter of Llyr. He is there described sitting on the rock of Harlech in Merioneth, with his courtiers around him, when he descried ships sailing rapidly towards the land: he sent some of his men down to meet them and to ascertain the business of those in them. It turned out to be the fleet of Matholwch, a king of Ireland, who was come to ask for Brân's sister Branwen to wife. Brân, without moving from his seat on the rock, conversed with Matholwch, and bade him land; after due deliberation the latter's request was granted, and he returned to Ireland with Branwen as his queen. Some years elapsed when, owing to court intrigues of uncertain details, Branwen was, for no fault of hers, disgraced and driven from the king to the kitchen, where she was badly used. At length she contrived to send word to her brother, who at once resolved to make an expedition to Ireland to avenge her wrongs. But instead of sailing across as the others did, Brân had to wade through the intervening waters, as no ship had ever been built of a size to receive this colossal king on board. As he approached the Irish shore, the swineherds of Erinn hastened to Matholwch's court with the strange story that they had seen a forest on the sea, and near it a great mountain with its spur flanked by two lakes: they added that both forest and mountain were in motion towards the land. Nobody could explain this until Branwen was summoned, and she told them that the trees were the masts of her countrymen's vessels, that the mountain they had seen must be her brother wading through shallow water, and that the mountain spur with the two lakes were his nose and eyes: she opined that his countenance betokened anger towards Erinn. Matholwch and his hosts hastened to place a river between them and the invaders. When the latter reached the stream, they found the bridge over it gone and the current impassable, until Brân laid himself across its bed and hurdles were placed on his body, so that his men passed over safely. He then got up and received ambassadors with offers of peace from Matholwch, which were rejected. But reconciliation was effected with Brân by the Irish paying him the compliment of building a house large enough for him, which was a novelty to him, as he had never before been able to enter a house. He had, however, a brother who was the genius of discord, and he succeeded in bringing on a quarrel between the hosts of Britain and Erinn, and in converting the spacious edifice into a slaughter-house, whence only seven of Brân's men escaped alive. The whole, it may be observed in passing, is the Celtic counterpart of the great carnage in the Nibelungen Lied; and the description of the waters[180] separating this country from Ireland suggests that the original story spoke of Hades rather than Erinn.

As to Brân himself, he rescued his sister, but received a poisoned arrow in his foot, whereupon he ordered the seven survivors of his army to cut off his head and to take it with them to their country. He told them that they would sit long at dinner at Harlech, and find the society of his Head as pleasant as it had ever been to them when it was on his body. From Harlech they would proceed, he said, to Gwales, the island now called Gresholm,[181] far off the coast of Pembrokeshire, and remain there feasting in the society of his Head so long as they did not open a certain door looking towards Cornwall. Once, however, they opened that door, it would be time for them to set out for London, and there in the White Hill to bury it with its face towards France; so long as the Head remained undisturbed in that position, this country would have nothing to fear from foreign invasion.

The Mabinogi relates how it happened to the seven men as Brân had foretold them; how they sat seven years at dinner at Harlech enjoying the friendly society of Brân's Head and the song of the Birds of Rhiannon, and how they enjoyed themselves eighty years in the island of Gwales. This part of the story takes its name from Brân's Urᵭawl Benn, that is to say, the Venerable or dignified Head. Time would fail me to discuss the probable identity of this Venerable Head with the Uthr Ben, the Wonderful Head, forming the subject of one of the most mythological poems ascribed to Taliessin;[182] and I will close these remarks with a brief mention of what would seem to be an echo of the Cernunnos myth in modern Wales. Lewis Morris, an antiquarian and patriot well known in the Principality in the last century, writing to a friend of his about his protégé the Welsh poet Goronwy Owen, complains of the uselessness of giving that genius of Bohemian proclivities any more money, as it seemed only to sink him deeper in difficulties, and he uses the words: 'It is surprising what confusion money will make. Is it any wonder that the devil should sit cross-legged in Ogo Maen Cymwd to guard the treasures there?'[183] Whether Lewis Morris had deliberately meant to represent the devil as a wellwisher of man need not be seriously discussed: what interests us at this point is the fact that the picture he projects of the devil reproduces exactly two characteristics of Cernunnos, in that he was supposed to sit cross-legged guarding the treasures in his cave. That a divinity like Cernunnos should end his career by being absorbed into the incongruous character of the devil, seems just what one might have expected.

With the aid of the Welsh instances, one is enabled to identify the Head in Irish literature likewise, where it is called that of Lomna, Finn's Fool, who is treated as an imbecile and as a poet or prophet: I allude to an article in Cormac's Glossary,[184] where verses ascribed to Lomna's Head are quoted. The story, strange and obscure as it is, may briefly be summarized thus: Finn and his men had set up their hunting-booth in Tethba, Anglicized Teffia, a district of considerable extent in the modern counties of Westmeath and Longford; and while Finn was busy with the chase, Lomna lurked about home, where he came one day across Cairbre, champion of the Luigni, with the Luignian woman who was Finn's wife in that district. She entreated Lomna not to tell Finn what he had seen; but unwilling to be a party to the disgracing of Finn, he wrote an Ogam couched in quaint metaphors, which Finn on his return did not fail to interpret. The result was that Cairbre, coming again at the suggestion of the Luignian lady when Finn was away, cut off Lomna's head and carried it away with him. Finn in the evening found a headless body in the booth and soon convinced himself that it was Lomna's; so the hounds were let loose, and Finn with his Fiann, as his men were called, tracked the murderer and his party to an empty house where they had been cooking fish on a stone, with Lomna's Head on a spike by the fire. The first charge cooked on the stone Cairbre divided, we are told, among his thrice nine followers; but he did not put a morsel of it in the mouth of Lomna's Head. Now this was a gess, 'prohibition,' to the Fiann, and the Head sang a very obscure verse about the fish. The second charge cooked on the stone was distributed as before, and the Head sang another verse, referring to the unjust division and foretelling the retribution that was quickly to follow. Cairbre then said to his men, 'Put out the Head, though it is an evil word for us.' No sooner had the order been obeyed than the Head outside uttered a third verse, and Finn with his Fiann arrived on the spot.

The association of poetry, prophecy and idiocy with one another is so thoroughly Celtic as to need no remark, and I would only point out that the offence taken by the Head at being refused a morsel of the meal which Cairbre and his men were making, seems to imply, that it was a custom to set apart some of one's food as an offering to the Celtic Dis, and that meals began with a religious ceremony of that nature.

Minor Divinities.

The foregoing gods and goddesses formed the leading figures in the fashionable circle, so to say, of the Gaulish pantheon; but some notice must now be taken of the crowd of minor gods and goddesses that also belonged to it. We must begin with the Genius Loci. Every Gaulish city, and British too probably, had its eponymous divinity, under whose protection it was supposed to be. The dedication of their cities was annually celebrated by the Gauls with libations made to the Genius of the place; and the practice lasted far into Christian times. The names of some of these Genii are known, such as Nemausus, Vesontio and Vasio, the tutelary divinities of Nînes, Besançon and Vaison respectively;[185] but the Allobroges have left no monuments of great interest in this respect. However, one from Geneva may be mentioned; it reads:[186] Deo Invicto, Genio Loci, Firmidius Severinus, mil(es) leg(ionis) viiiae Aug(ustae) P(iae), &c. The name of the god is not given, and all we learn from the inscription is that he was assimilated with Deus Sol Invictus or Mithras, whose worship was introduced by the Romans, as inscriptions in various places in the Allobrogic land clearly prove.

Next come the mother goddesses, who are usually called, in Gaulish inscriptions written in Latin, Matrae, Matres or Matronae: the dative is the only case that occurs in those of the Allobroges, and it is found to be matris or matrabas,[187] for which the pure Gaulish form would have been mâtrebo, as in a vernacular inscription at Nîmes.[188] As figured in ancient bas-reliefs, the Mothers take the form of three young women of a benevolent countenance and clad in long robes. They are mostly represented in a sitting posture, with fruit on their laps or, occasionally, an infant on their knees; but an uninscribed altar in the museum of Vienne shows the Mother in the middle sitting with a basket full of fruit on her lap, while her two sisters stand by her in long robes which cover their heads; and a monument found at Metz represents the three standing, and holding in their hands fruit or flowers. To come back to the land of the Allobroges, one is led to suppose, from the number of inscriptions and sculptures in honour of the Mothers, that their worship was far the most popular there, and the same remark would probably apply to the rest of Gaul.

Usually, but not always, they have the official title of Augustae given them, as in the following inscription found at Sainte-Colombe, near Lyons: Matris Augustis, C. Titius Sedulus ex voto.[189] And a remarkable specimen on a piece of stone at the church of Aoste, near La Tour-du-Pin in the Isère, reads:[190] Matris Aug(ustis) ex stipe annua denariorum xxxv et d(onis). . . . Here it may be observed that an offertory of only thirty-five denarii in the course of a whole year is pretty clear evidence that the faithful who practised the cult of the Mothers at Aoste were all poor people, and that the service was of the simplest and most rustic nature. One more of these inscriptions deserves to be mentioned, namely, that found at Grenoble, and reading, so far as it is still legible:[191] Matris Nemetiali(bus) Lucretia Q. Lib(erta) . . . . After the analogy of other instances, Nemetiali is taken to be an abbreviation for Nemetialibus, and this last has further been supposed to be a sort of Gallo-Roman equivalent of the official Augustis; but it is much more likely to be a topical epithet referring to some spot where they had a shrine: compare Matrebo Namausicabo, 'to the Mothers of Nîmes,' and Matribus Treveris, 'to the Mothers of Treves.' But if this should be thought an unsatisfactory explanation, several others might be easily suggested: for instance, Nemetialibus might mean that the goddesses were referred to as worshipped in νεμητα or groves, as it was not unusual for sacred spots to have a grove of trees close by the shrine of the presiding divinity; nor was it otherwise in the case of the Mothers, as may be seen from an inscription found at Monterberg, near Xanten, reading: Matribus Quadruburg. et Genio Loci Sep. Flavius Severus Vet. Leg. X. G. P. F. V. V. Templum cum Arboribus constituit.[192] But as νεμητον which we have already met with in the Gaulish inscription found at Vaison, may be regarded as meaning not simply a grove, but a sacred grove, and connected with the Welsh word nyfed, supposed to mean 'sanctity or purity,' it would perhaps be right to render Nemetialibus by Sanctis, and to compare the Sanctis Virginibus of an inscription in a cartouche on a broken pillar at Saint-Romain-en-Gal, near Lyons.[193] These Virgins probably belonged to much the same class of divinities as the Matrae: the probable reading of the whole inscription, which is in a bad state, is said to be: Sanctis Virginibus, Sacrum Avitus (et) Campana posuerunt.

Christianity failed to put an end to the belief in these divine Mothers and Virgins: it was continued in connection with benignant fairies and the Madonna, whence a certain number of the churches known by the name of Notre-Dame, built on spots where legend asserts images of the Virgin to have been miraculously discovered. These heathen statues are usually of wood, which has turned black in the ground, whence the chapels of the Black Madonnas, such as the one in the commune of La Tronche, near Grenoble. The story there goes that the statue of the Virgin, consisting of black wood, was discovered in a vineyard; and, though the original has been since lost and superseded by one of blackened stone, the sanctuary of the Vierge Noire is very popular, pilgrimages being made to it on Whit-Monday by persons, especially of the fair sex, who wish to be married; if the Virgin is favourable, the desired marriage takes place, we are told, within the year.[194]

Besides Matres and Virgines, Gaulish paganism had also its Dominae, who cannot have been very different from them: thus a tablet found at Saint-Innocent, near Chambéry in Savoie, reads:[195] Dominis exs voto s(olvit) l(ibens) m(erito) M. Carminius Magnus pro salute sua et suorum. Lastly, a Gaulish word which may possibly be the equivalent of this Dominis is used in an interesting inscription found at Aix-les-Bains, in Piedmont, which reads as follows:[196] Comedovis Augustis M. Helvius Severi Fil(ius) Iuventius ex voto. No other monument in honour of the Comedovae is known except one near Cologne, on which the dative has been read both Comedovibus and Comedonibus.[197] In either case we have a common stem comed-, which is also that of the Old Irish word coimdiu or comdiu, genitive comdeth, which is found mostly applied to God: in modern Irish it is superseded by tighearna, 'lord.' It has been analysed con-med-, in which med would be of the same origin as Welsh 'meᵭiant,' 'power, authority, ownership:' we have the same root probably in the English word mete, 'to measure;' and one may go further and point out that the stem mediot, of the Irish comdiu, comdeth, is probably to be found bodily in the Anglo-Saxon metod, which applied only to God, and in its Norse equivalent 'mjötuᵭr,' as when in the Volospá mention is made of 'a glorious judge beneath the earth,'[198] supposed to be Heimdal. These facts suggest that the word Comedovae was applied to a class of Matrae or Dominae, who, in their capacity of guardians of the weak against the strong, were supposed to discharge to some extent the functions of a judge or dispenser of justice by avenging his crimes on the wrong-doer; but it would probably be a mistake to suppose them to have partaken to any great extent of the character of the Greek Erinnyes.

As to these last, however, it may be gathered from later indications that vulgar imagination peopled all Celtic lands at the early time we are speaking of, with an indefinite number of hurtful and malevolent spirits, goblins and ogresses of all kinds, whose cult of terror seldom probably attained to the dignity of monumental record. But a characteristic exception occurs in this country in the case of an inscription found at Benwell, near Newcastle on Tyne: its brevity is remarkable, as the whole consists of the two words Lamiis Tribus,[199] 'To the Witches Three' The devotee, who did not wish his name known, may have been a soldier from the Continent; but the three Lamiae were doubtless as British as the Three Witches in Shakspear's Macbeth.

It has often been noticed that all these divinities, whether friendly or unfriendly to man, usually muster in threes; but they probably resembled the Genius Loci in being local and attached to particular spots, and it would therefore not be always easy to draw a clear line of demarcation between them and the other gods and goddesses associated with the salient features of a Celtic landscape. The number of these minor divinities was legion; and, without attempting to draw a hard and fast line between them and the greater divinities, who also lent themselves to localization, one may say that among the former must be included the spirits of particular forests, mountain tops, rocks, lakes, rivers, river sources, and all springs of water, which have in later times been treated as holywells, whether in France or the other lands inhabited by the Celts. It has been supposed, and not without reason, that these landscape divinities re-acted powerfully on the popular imagination in which they had their existence, by imparting to the physical surroundings of the Celt the charm of a weird and unformulated poetry. But what race was it that gave the Celtic landscape of antiquity its population of spirits? The Celtic invaders of Aryan stock brought their gods with them to the lands they conquered just as much as the ancestors of the English brought with them theirs to the Christian land of Roman Britain; and the former continued to be in the main the great figures of the Celtic pantheon until that rude edifice crumbled to dust under the attacks of Christianity; but as to the innumerable divinities attached, so to say, to the soil, the great majority of them were very possibly the creations of the peoples here before the Celts. In any case, it is curious to observe that, while Christian missionaries appear to have made comparatively short work of the greater Celtic gods of Aryan origin, the Church fulminated in vain against the humble worship of wells and stocks and stones. That cult required no well-defined and costly priesthood which could be overturned once for all, and, a little modified, it thrives in some Celtic lands to this day. All that the Church could do was to ignore it for a time, and ultimately to assimilate it: to effect its annihilation has always been beyond her power. We have a good parallel in modern Greece, of which it has been well said,[200] that 'the high gods of the divine race of Achilles and Agamemnon are forgotten, but the descendants of the Penestæ, the villeins of Thessaly, still dread the beings of the popular creed, the Nereids, the Cyclopes, and the Lamia.' The greater divinities of the Celts were undoubtedly Aryan; but to what extent the motley mob of lesser gods and goddesses were likewise Aryan is a question that must remain unanswered, until, at any rate, the ethnologist, the archæologist and the historian have succeeded in placing before us in a clearer light than has yet been done, the rôle played by those ancient peoples who occupied the west of Europe before the Celts first beheld the Atlantic or approached the sunny shores of the Mediterranean.


  1. Bellum Gallicum (ed. Holder), vi. 17.
  2. For the substance of these remarks I am indebted to an excellent article by M. Florian Vallentin, entitled, Les Dieux de la Cité des Allobroges, in the Revue Celtique, Vol. iv. 1—36, to which I shall have frequently to refer in this lecture.
  3. Vallentin, ibid. p. 1; see also Desjardins, Géographie historique et administrative de la Gaule romaine, ii. 351.
  4. Vallentin, Rev. Celtique, iv. 1, 2.
  5. Inscriptions antiques et du Moyen Age de Vienne en Dauphiné; consisting of six octavo volumes of letter-press description of them, supplemented by a quarto one of plates, published at Vienne in the year 1875.
  6. Allmer, iij. 112; Vallentin, Rev. Celtique, iv. 17.
  7. Mommsen, Inscriptiones Helveticae (in Vol. x. of the Mittheilungen der Antiq. Gesellschaft in Zürich), No. 215; Rev. Celt. iv. loc. cit.
  8. Brambach's Corpus Insc. Rhenanarum, No. 1591.
  9. Allmer, iii. 191, pi. 38-8; Rev. Celt. iv. 16. There seems to be some doubt as to whether Magniaco or Macniaco is the correct reading: Allmer gives both without remarking on the discrepancy.
  10. The Gauls, like the modern Celts, had no objection to compound terms, and they even used foreign elements in such place-names as Augustonemetum, the grove of Augustus; Caesarodunum, Caesar's fortress; and Juliomagus, the field of Julius. Some of their personal names were quite as long: witness Conconnetodumnos, Veriugodumnos and Vercassivellaunos. These and the like must have seemed cumbrous to the Romans; and Englishmen of the present day profess to be amused with German compound terms, forgetting that they are usually the shortest way of expressing what is meant, and that few languages form compounds more readily or complicately than their own, though the longer terms are never written as single words: take, for example, such instances as 'university examinations,' 'university examination-papers,' 'London, Chatham and Dover Railway,' or 'London, Chatham and Dover Railway Company.' It is only an accident—doubtless an inseparable accident of perversity—that English grammarians usually conceal the fact of the composition.
  11. See them enumerated by M. Vallentin in the Rev. Celt. iv. 15.
  12. Esquisse de la Religion des Gaulois (Paris, 1879), p. 10.
  13. A somewhat shorter one was given by M. Mowat in the Rev. Archéologique (1875), Vol. xxix. p. 34, where he gives a reason for connecting the place-name Montmartre with the god, a view also taken by M. Gaidoz.
  14. Rev. Celt. iv. 15.
  15. Jollois, Memoires sur quelques Antiquités remarquables du Département des Vosges (Paris, 1843), p. 126, &c.
  16. Rev. Celt. iv. 15 and ii. 426; Bulletin Monumental, 1875, p. 557, et seq.
  17. See also Mowat in the Rev. Arch. (1875), Vol. xxix. 31.
  18. Rev. Celt. ii. 426, iv. 15; Rev. des Soc. savantes (1875), Vol. i. p. 249.
  19. Brambach, Nos. 256, 257, 593, 1741, 2029 add. p. xxvii.
  20. Kuhn's Beitræge zur vergleichenden Sprachforschung, iij. 169; Brambach, No. 835; Rev. Arch. (1875), Vol. xxx. pp. 367, 368, where M. Mowat corrects Brambach's Vasso·caleti into Vassocaleti. In the same article, p. 361, he gives facsimiles of the readings of the corresponding form in the chief manuscripts of Gregory's text.
  21. Gregorii Turonensis Opera, Historia Francorum (contained in Monumenta Germaniæ Historica), Lib. i. c. 32, where the reading preferred by the editors begins with Veniens vero Arvernus, &c., but A 1 reads Arvernos.
  22. Even those who preferred doing so would have to explain Vasso Galate as meaning the Gaulish temple, and to refer it probably to the same edifice.
  23. Much conjecture has been wasted on this term, especially by writers aware only of a Welsh word gwas, meaning a young man or servant, Gaulish vassos (as in Dagovassus), and not of gwas, meaning a palace or mansion, which alone is the one here in point.
  24. This is according to a rule still obtaining in Welsh, as when we say Ivan Hirnant, 'Evan of Long-brook,' or Tudur Penllyn, 'Tudor of Penllyn,' in both of which the place-name is to be construed as a genitive; and we have an instance from a time before the case-endings were dropped, in a bilingual inscription from Brecknockshire, which reads Maccutreni Saliciduni, '(the Stone of) Maccutreni of Salicidunon' (Rhys, Lectures on Welsh Phil. p. 382). Then as to the compound Vasso-calet, one has to compare the Welsh treatment of permanent epithets. Thus we say Maelgwn Fychan, 'M. Vaughan or M. the Little,' while a little Maelgwn, to whom the adjective was not constantly applied, would be Maelgwn bychan, 'little Maelgwn.' Put back into an early form, the latter would be Maglocunos biccanos, while the former would be Maglocuno-biccanos; and it is in this way that I would explain the Gaulish Vasso-caleti as a compound in the genitive case. Compare the Irish genitive na Crǽb-rúadi, in the Bk. of the Dun, 99b: it was the name of the king of Ulster's palace, and literally meant the Red Branch, a designation, however, of uncertain connotation. One may also probably compare the Ogmic genitive Neta-Ttrenalugos, with tt for later th, and a neta which is in my opinion not a genitive.
  25. Rev. des Soc. sav. Vol. i. (1875), p. 250; Rev. Celt. ii. 426.
  26. Cormac's Glossary, translated by O'Donovan and edited by Stokes (Calcutta, 1868), p. 40; Meyer's Ventry Harbour (Oxford, 1885), x. 87; O'Curry's Manners and Customs of the Ancient Irish, i. pp. dcxxxvij, dcxxxix.
  27. The Berlin Corpus Inscr. Latinarum, Vol. vii. (Inscr. Britanniae Latinae, edited by Hübner), No. 271.
  28. In fact, the god's equipment shows that a determined effort had been made to get him up in the classical way.
  29. I am not quite sure that I comprehend this allusion to the lily; but here is the original for those who may object to being led astray: καὶ οἱ ἀγορηταὶ τῶν Τρώων τὴν ὄπα τὴν λειριόεσσαν ἀφιᾶσιν εὐανθῆ τινα· λείρια γὰρ καλεῖται, εἴ γε μέμνημαι, τὰ ἄνθη. The whole prolalia is No. 7 (pp. 23—25) in Bekker's edition, and No. 55 (pp. 598—600) in Dindorf's; extracts from it will also be found in Zeuss's Grammatica Celtica, edited by Ebel, pp. 1, 2.
  30. Rhys's Lectures, pp. 293—295.
  31. Mr. M. Atkinson (quoting from the Irish MS. called the Book of Ballymote), in the Kilkenny Journal of the Royal Hist. and Arch. Ass. of Ireland, for 1874, p. 207; see also my Lectures on W. Phil. p. 293.
  32. O'Donovan, Irish Grammar, p. xlviij; also the Rev. Celt. vii. 369-74, where the true nature of a large part of the Ogmic jargon has been explained for the first time by Thurneysen.
  33. Gram. Celtica, p. 2; Rhys, Lectures, p. 298.
  34. Böhtlingk and Roth's Sanskrit Dictionary, s. v. ajma, ajman.
  35. The Kilkenny Journal for 1874, p. 229, and my Lectures on Welsh Phil. p. 293, where I have rendered Grian-ainech—less correctly, as I now think—'sun-faced.'
  36. Rev. Celt, iv. 25.
  37. Hübner, Nos. 218, 332, 1345.
  38. See Skene's Four Ancient Books of Wales, ii. 174.
  39. The Berlin Corpus, iii. Nos. 1130, 1132, 1133, 1136, 1137, 1138.
  40. Brambach, No. 1915.
  41. Rev. Celt. iv. 134; Mém. de la Soc. royale des Antiquaires de France (1823), v. p. xxii.
  42. Rev. Celt. iv. 144.
  43. The Berlin Corpus, iii. No. 5861.
  44. Greppo's Eaux thermales ou minérales de la Gaule (Paris, 1846), p. 29.
  45. Vallentin, Rev. Celt. iv. 446.
  46. Ibid. iv. 6, 9.
  47. Greppo, p. 56.
  48. Eumenii Panegyricus Constantino Dietus, xxi, xxij (in Migne's Patrologia, viij; sec col. 637-8); Greppo, pp. 51, 52; Rev. Celt. iv. 144.
  49. Rev. Celt. iv. 137—139.
  50. See The Annals of the Kingdom of Ireland [compiled, in the 17th century, by the Four Masters, and edited by O'Donovan (second edition, Dublin, 1856), a work which will briefly be referred to as the 'Four Masters' in the rest of this volume], A.M. 4003, 4019, 4020, 4169, 4178.
  51. The Text of the Mabinogion, &c., from the Red Book of Hergest, edited by Rhys & Evans (Oxford, 1887), pp. 124, 131-2. [This text will hereafter be referred to as R. B. Mab.] Guest's Mabinogion, ij. 287-8, 300-1.
  52. R. B. Mab. p. 141; Guest, ii. 315.
  53. Brambach., No. 1529.
  54. One of the most recent writers on this subject is Wilhelm Müller, in his Mythologie der deutschen Heldensage (Heilbronn, 1886), pp. 148—189; and a succinct account of the original literature embodying the Dietrich legend will be found in Karl Meyer's Dietrichssage (Basle, 1868), pp. 4—9.
  55. W. Müller, p. 186; Jahrbücher des Vereins von Alterthumsfreunden im Rheinlande, i. 12—21, xiij. 1.
  56. See Hodgkin's Italy and her Invaders (Oxford, 1885), iij. 341.
  57. Rev. Celt. iv. 10; Mommsen, Insc. Helvet. No. 70; Allmer, iij. 255.
  58. Brambach, No. 1588.
  59. De Bonstetten, Receuil d'Antiquités suisses (Berne, 1855), pp. 35, 37.
  60. Rev. Celt. iv. 11; Rev. Archéologique (1852), ix. 315.
  61. Rev. Celt. iv. 11; Monnier, Annuaire du Jura for 1852, plate 1, which I have not been able to consult.
  62. Mém. de la Soc. des Ant. de France (1850), xx. 58-9.
  63. Rhys, Lectures, p. 395; Stokes, Celtic Declension, which appeared first in Vol. xi. of Bezzenberger's Beitræge (Göttingen, 1886), p. 87.
  64. Gruter, lviii. 5.
  65. Nevertheless, the name is not given by Brash in his book on The Ogam inscribed Monuments of the Gaedhil; but by correctly reading Brash's copies I had detected it in the case of the Stradbally inscription (Brash, p. 254, pl. xxxv), and of one of those at Ardmore (Brash, p. 247). In 1883 I had the pleasure of seeing, by inspection of the stones, that my readings were correct, and also of finding Netas(egam)onas in an inscription at Seskinan (Brash, p. 262).
  66. Nia Segamain in the Book of Fenagh, edited by Prof. Hennessy (Dublin, 1875), p. 29, and simply Nia at p. 56; the Four Masters, A.M. 4881, 4887, 4990, write Nia Sedhamain (dative) and Niadh Sedhamain (genitive). The older and more correct forms would be, genitive Segamon or Segaman, dative Segamain, unless there was an optional O-stem.
  67. Rev. Celt. iv. 12; Congrès Arch. de France, 1873, p. 215. Can Bouhy and Bolvinnus be of the same origin?
  68. Rev. Celt. iv. 12; Mém. de la Soc. des Ant. de France (1850), xx. 59.
  69. Rev. Celt. iv. 23 ; Allmer, iii. 243.
  70. Rev. Celt. ib.; Allmer, ii. 345.
  71. Rev. Celt. iv. 24.
  72. Rev. Celt. iv. 25.
  73. Hübner, Nos. 286, 643, 886, 914, 977.
  74. Gaidoz, Esquisse, 10.
  75. Hübner, Nos. 294, 333, 369, 745, 873, 934, 935. Belatucadrus ends with a word to be identified with the Welsh cadr; 'powerful, strong, robust;' Breton caer, formerly cazr, 'beautiful, fine, magnificent:' so the whole word means fine or powerful at the kind of action indicated by the vocable belatu, which has the appearance of being a verbal noun. We have the stem bel (mutated into fel and pronounced vel) in the Welsh word for war, namely, rhyfel; it is also added to oer, 'cold,' to make oerfel, 'cold weather,' or cold as productive of inconvenience and harm. Again, we have it in ufel, 'fire or conflagration,' Irish óibell, óibel, which meant a spark, fire or heat, and was applied, for instance, to the summer heat that drives cattle to stand in pools; the other element in these words is the Celtic reflex of the first syllable of the Greek ἄνω or of the Latin uro, 'I burn.' Irish supplies us with a strung verb from the stem bel, as in bebla (for be-bela), 'mortuus est,' atbail, 'interit,' atbél, 'peribo.' A corresponding Welsh compound has yielded a derivative adfeilio, 'to decay or fall into ruins;' but the Irish verb had as its base bel, meaning 'to die,' while belatu implies a derivative verb from a theme bela, associated probably with a modified meaning, namely, the causative one of killing or slaying; and an instance of it occurs in Welsh in a poem in the Book of Taliessin, where reference is made to the cattle of the Egyptians killed by the fifth plague or the grievous murrain spoken of in the Book of Exodus, ix. 1—7. See Skene, ii. 171, where the form used is belsit, which would seem to mean 'had been killed.' Having found a strong verb bel, we ought to be able to identify it in some of the kindred languages: now the Aryan combination gv becomes b in Celtic, while in the Teutonic languages it would be hardened into cw or qu; so we look in them for a verb beginning with quel or cwel to correspond to our Celtic bel, and we readily find it, without going out of this country, in the Anglo-Saxon verb cwelan, 'to die or perish,' from which was formed a causative cwellan or cwelian, 'to slay or cause to perish,' represented by the modern English verb to kill.
  76. Brambach, 164.
  77. At Rome: see the Berlin Corpus, vi. No. 46; and a Camulo Viromanduo is reported from Auvergne in the Rev. des Soc. sav. (1875), i. 251.
  78. Hübner, 1103.
  79. What is the meaning of the word in the post-Roman personal names Camelorigi from Pembrokeshire, and Camuloris, Camulorigho, from Anglesey? For some notes on them, see my Lectures, pp. 364, 400. The name Camulogenus, which Caesar (Bell. Gall. vii. 57, 59, 62) gives the defender of Paris against the legions of Rome, would mean the descendant of Camulos, and similarly Camulognata.
  80. On the difficulties of this etymology, see Kluge's Etym. Wörterbuch des deutschen Sprache, s. v. Himmel.
  81. Max Müller's Lectures on the Science of Languages8, ij. 473; Rig-Veda, iv. 17, 4.
  82. Hübner, 61.
  83. Orelli-Henzen, Vol. iij. No. 5867; J. de Wal, Mythologiae Septentrionalis Monumenta epigraphica Latina, No. ccxcij.; L'Institut for 1841, p. 160.
  84. Hübner, 36.
  85. Esquisse, p. 10; but this is not certain, and the name seems to be the same that was meant by Niámán in the Bk. of Leinster, 81b, printed in O'Curry's Manners, &c., iij. 418-9. The former, I may say in passing, is a MS. of the 12th century, and my reference is to the lithographed facsimile published with an introduction by Prof. Atkinson, Dublin, 1880.
  86. Rev. Celt. iv. 19.
  87. Caesar, Bell. Gall. ii. 23.
  88. Hübner, No. 84. An inscription found at York, and another at old Carlisle, suggest the respective spellings Totati (Hübner, Ephemeris Epigraphica, iii. 313, No. 181) and Tutati Cocidio (ibid. p. 128; Hübner, No. 335): they are difficult to read, but Totati is countenanced by the name Totatigens borne by a Gaulish soldier in the Cohortes Vigiles at Rome (Rev. Celt. iii. 272; Berlin Corpus, vi. 2407), while Tutati must be left doubtful, though not only Toutati, but also Tōtati and Tūtati seem to be perfectly legitimate forms. Here it may also be mentioned that Toutati derives some confirmation from a monument at the other extreme of the Celtic world, namely, one found at Seckau in Styria: see the Berlin Corpus, iii. 5320; but M. Mowat, in the Bull. Epigraphique de la Gaule, i. (1881), 123, reads, not Toutati, but Tioutati, which has a suspicious look about it. M. d'Arbois de Jubainville is inclined to contest M. Mowat's reading: see his Cycle Mythologique irlandais et la Mythologie celtique, p. 378.
  89. Stokes in Kuhn's Beitræge, i. 451, ii. 107; Becker, ibid. iii. 162.
  90. Also a parallel of a different etymological origin in the Old Norse fylkir, a poetic word for king, derived from fólk; and the derivation of the word king itself, Anglo-Saxon cyning, is in point, though it involves several difficulties. See Kluge, s. v. König.
  91. The singular of this word would be the Teutonem, which Holder has preferred, in his recent edition of the Germania, to the more usual Tuisconem or Tuistonem.
  92. Toutates, which the scanning of Lucan's verse would make into Toutātes, was apparently formed in the same way as the Gaulish Dunates and Dumiates already cited, to which may be added Baginates, to be mentioned presently.
  93. The original note will be found in Usener's Lucani Commenta Bernensia (Leipsic, 1869), p. 32: 'Hesum Mercurium credunt, si quidem a mercatoribus colitur, et praesidem bellorum et caelestium deorum maximum Taranin Iouem adsuetum olim humanis placari capitibus, nunc uero gaudere pecorum.'
  94. See the first of M. Gaidoz's Études de Mythologie gauloise (Paris, 1886), pp. 5-6 and plate; also M. Rochetin's article in the Mém. de l'Acad. de Vaucluse, 1883, pp. 184—189, which I have not been able to consult.
  95. Rev. Celt. iv. 21; Allmer, iij. 197.
  96. Baginates admits also of being derived from the same root as the Latin word fagus, and in that case one might compare the Dodonian Ζεὺς Φηγός or Φηγοναῖος: see Overbeck's Griechische Kunstmyfliologie, i. 4. What did the Phrygian epithet of Ζεύς Βαγαῖος mean?
  97. Rev. Celt. iv. 13; Allmer, ij. 454.
  98. Rev. Celt. iv. 14; Mommsen, Insc. Helv. No. 140.
  99. Hübner, Ephemeris Epigraphica, iij. 313 (No. 181); Rev, Celt. iv. 446.
  100. Gaidoz, Études, p. 105; Jahrbücher des Vereins von Alterthumsfreunden im Rheinlande (1882), lxxiv. p. 189; Bull. épigr. de la Gaule, iij. (1883), p. 154; iv. (1884), p. 200.
  101. Études, p. 96.
  102. Ib. pp. 93, 96.
  103. Études, p. 98.
  104. Ib. p. 90.
  105. Ib. p. 91.
  106. Berlin Corpus, iii. No. 2804; Rev. Celt. v. 385.
  107. Brambach, 1589, 1812; Rev. Celt. v. 386.
  108. Allmer, ij. p. 426; Rev. Celt. iv. 21, v. 383, where it is given as I. O. M. Fulguri, &c.
  109. Hübner, No. 168. Since writing the above I have found that M. Gaidoz, Études de Myth. gaul. p. 97, suggests the same idea as I do as to the nationality of the inscription: I have again examined the stone, and I am obliged to admit that the reading of the god's name is doubtful; but one thing is certain, namely, that it was never Tarano, as some will have it: that is out of the question. The alternative reading which the present state of the stone would suggest would be something like Innaro with nn conjoint.
  110. Compare the A.-Saxon reference to the thunder 'with the fiery axe' (mid đǽre fýrenan æcxe) in the Dialogue of Salomon and Saturn, ed. Kenable (London, 1848), p. 118, also 177.
  111. By Vigfusson and Powell in their Corpus Poeticum Boreale, ij. 463.
  112. Jordanis de Origine Actibusque Getarum (ed. Holder), cap. 13, p. 18.
  113. Vigfusson & Powell, Corpus Poet. Boreale, ij. 464; Cleasby & Vigfusson, Icelandic-English Dic. s. v. áss, which is the Norse form.
  114. Hübner, No. 1334,61; Mommsen, Inscr. Helv. No. 80.
  115. In living Celtic the s of Esugen- must, according to analogy, be dropped, and Eugen- becomes in Irish Eogan, now written Eoghan, owing to the shifting of the accent forward leaving the vowel of the last syllable neutral, while in early Brythonic the reverse was more nearly the case, and Eugen- was made the basis of a derivative Eugenjos, whence the Welsh forms Eugein, Euein and Ywein, of which the colloquial form was Owein, now reduced to Owen: compare Welsh bywyd, 'life,' colloquially pronounced bowyd, and the like. The oldest spelling appears to have been Eugein, which was Latinized as Eugenius, just as Anglo-Irish Eoghans call themselves Eugenes.
  116. Bull. Épigr. i. 62—68.
  117. Figured and described by M. Anatole de Barthélemy, in the Musée Archéologique, ij. (1877), p. 8; and in Mélusine, col. 353: see also Gaidoz's Études, p. 88.
  118. Gaidoz, Esquisse, pp. i, 11; Cerquand, Rev. Celt. v. 386.
  119. By A. de Barthélemy, Rev. Celt. i. 1—8, where the attempt is made to prove the personage meant to have been Dis Pater; but that is no longer the way to look at the question, since M. Mowat has succeeded in showing that the infernal deity is to be identified with the Gaulish Cernunnos, as will be shown later.
  120. According to M. A. de Barthélemy, the dog lias three heads, so that he treats it as a Cerberus (Rev. Celt. i. 3), and the same view is adopted by M. Flouest (Rev. Arch. 1885, v. 20). An engraving of the altar will be found in the Rev. Arch. 1879, xxxvij. pl. xii, and another in a previous volume, 1854, where M. Chardin also gives the dog three heads, p. 310.
  121. Bull, épig. de la Gaule, i. 65; Berlin Corpus, iij. 1152, where Mommsen suggests that the second word was meant to be domestico; but the reading of the letters originally written is difficult.
  122. Hübner, No. 642.
  123. Rev. Celt. i. 1—8.
  124. Rev. Arch. v. (1885), 7—30.
  125. I owe my information to the kindness of Dr. Stokes, who states that the inscription I allude to in the text was discovered at Orgon, in the Bouches du Rhône. The Brythonic word for thunder is taran, masculine in Breton and feminine in Welsh. This taran does not correspond declensionally to Taranis, but it may either to taranus of the U declension, or else to forms of the declension, such as taranos (masc.) or taranon (neut.), with which the Goidelic ones agree, namely, Irish torunn, 'thunder;' Sc. Gaelic torrunn, the same. These last languages had also toirn, or tairn, of a different declension, of which more anon. The rest of the difference between toirn and torunn is paralleled by the Irish iarn, 'iron,' which takes also the form of iarann, of the same meaning.
  126. See also Cerquand, Rev. Celt. v. 381-8; Mowat, Bull. Épig. i. 123-9.
  127. I examined it in 1883, but could not feel quite certain whether Toranias or Turanias was the better reading, though I was inclined to the former.
  128. By O'Huidhrin: see The Topographical Poems, edited by O'Donovan (Dublin, 1862), p. 102, and note lxiv (555).
  129. There was also a poet called Torna supposed by some to have lived in the 4th century, and it has become usual to trace the Ui Torna to him as their ancestor. This is probably an error dating from the time when a nominative Torna would be Torna also in the genitive; the former would presuppose an early Toranjo-s, an adjectival form of the same origin as the words here in question, but parallel with such Latin formations as Jovius, Martius, Veneria, and the like. However, it would matter little here if one were forced to suppose some of the Torna families descended from the poet alluded to; the rest may still be regarded as deriving their name from Torna = the genitive of Toranjas.
  130. O'Reilly's Irish-Eng. Dict. s. v.; see also Foley's English-Irish Dict. s. v. thunder.
  131. Mommsen, Inscr. Helv. No. 215, where the inscription is read: Deae Artioni Licinia Sabinilla.
  132. Orelli, 1431, 1969. The two entries seem, owing to an oversight of the editor's, to represent one and the same inscription.
  133. Hübner, No. 97. Belismius is the reading adopted by Hübner, but it is not at all certain: see Lee's Isca Silurum, pp. 19—21, pl. viii. 1, where he reads Belisimnus. What one would expect is Belisamius, Belisemius or even Belisimius.
  134. Three Irish Glossaries, ed. by Stokes (London, 1862), p. 8, and O'Donovan's translation of Cormac's Glossary, ed. by the same scholar (Calcutta, 1868), p. 23.
  135. Stokes's Three Ir. Gloss, p. xxxiii ; M. d'A. de Jubainville's Cycle Mythol. p. 146; Rhys's Celt. Brit. p. 282.
  136. One comes from Doncaster, and one from the neighbourhood of Leeds; the other two belong to the line of the Roman Wall in Cumberland, and to Middleby in Scotland respectively: they are numbered 200, 203, 875 and 1062 in Hübner's volume of the Berlin Corpus. The last mentioned is, unfortunately, the only one which preserves the name of the goddess in full. Thurneysen, in Kuhn's Zeitschrift, xxxviij. 146, equates the name with the Sanskrit 'Bṛihatî,' 'die höhe.'
  137. De Belloguet's Ethnogénie gauloise2, i. 289; Stokes' Celtic Declension (Göttingen, 1886), p. 67.
  138. For instance, in the British Museum MS., Harl. 5280, fol. 68a.
  139. See Sullivan's note, p. clxxi of the introductory volume to O'Curry's Manners, &c.; O'Curry's MS. Mat. p. 46; Windisch's Irische Texte, p. 266 (§28). A Bríg Briuguid is alluded to in the Senchus Mór. i. 144.
  140. See Celt. Britain2, p. 282.
  141. Bull. Épigr. de la Gaule, i. 111—116.
  142. Ibid. p. 113; Orelli-Henzen, No. 6087; B. Corpus, Vol. iii. p. 926.
  143. Hesychius, κάρνον · τὴν σάλπιγγα Γαλάται; and Eustathius (Leipsic, 1829), ad Homeri II. 1139, 57, κάρνυξ.
  144. With the Gaulish divinities of this kind M. d'Arbois de Jubainville would compare, Cycle Myth. p. 384, an Irish Buarainech, which he renders 'à la figure de vache:' the person so called is said to have been the father of Balor.
  145. Bull. Épigr. i. 111-2.
  146. L'Autel de Saintes et les Triades gauloises is the title of an able and copiously illustrated account of the most important monuments representing the Gaulish Pluto, by the well-known keeper of the museum at Saint-Germain, M. Alex. Bertrand, in the Revue Archéologique for 1880: I refer here more especially to the offprint, published in Paris in 1880: see pp. 1, 7, 38; also the Rev. Arch, for 1882, xliij, p. 322, and plate ix.
  147. Bull. Épigr. ibid.; also Les Antiquités d'Entrain, by M. de Villefosse, which I have not been able to consult.
  148. Bull. Épigr. ibid.; Bertrand, pp. 7, 8.
  149. Bertrand, p. 7, pl. xi. Why M. Mowat regards the contents of the bag which the god empties as pieces of money (Bull. Épigr. i. 112) I do not exactly understand.
  150. Bull. Épigr. ibid.
  151. It is worth mentioning, as bearing on the question of the distribution of the statues of the god, that a mutilated one of him was discovered in the department of the Puy de Dôme in the year 1833: if not again lost, it should be now in the museum at Clermont. I owe this information to a notice by M. Gaidoz in the Rev. Arch. for 1880, iv. 299—301. See also the Rev. Arch. for 1882, xliij. 125, where the wide distribution of the tricephal has induced M. Mowat to declare for the improbable hypothesis, that it was after all but the Roman Janus more or less completely naturalized in Gaul.
  152. Mowat, Bull. Épigr. de la Gaule, i. 114-5, where he gives, besides other authorities, Diod. Siculus, v. 27; Pliny, Hist. Nat. xxxiij. 10, and others; also sundry coins and inscriptions.
  153. Strabo, iv. 4, 3; Diod. Sic. v. 28.
  154. Bull. Épigr. de la Gaule, i. 115; Diod. Sic. v. 30.
  155. In their Corpus Poeticum Boreale, ij. 465.
  156. Corpus Poet. Bor. ij. 114.
  157. Bertrand, pp. 10, 28—31.
  158. It is the one with the name and figure of Esus on its northern face, while its principal face looking towards the east hears the figure and name of Iovis, and the southern one those of Volcanvs: see the Bull. Épigr. i. pp. 60, 61, 68.
  159. See Stokes's reading in the Academy for Sept. 25, 1886, p. 210a.
  160. This also is a discovery of M. Mowat's, ibid. pp. 68—70.
  161. Bertrand, pp. 20, 33-9.
  162. Glück's Keltischen Namen, p. 97.
  163. The consonantal declension was always liable to be replaced, so we have Donand and Danann used in the nominative, whence a new genitive, Danainne, was sometimes made. See O'Donovan's note, Four Masters, A.M. 3450 (i. p. 23), A.D. 1124 (ij. p. 1020).
  164. The reference is to the Triads published in three versions in the Myvyrian Archaiology of Wales (London, 1801), Vol. ij. pp. 1—22. Version ij. is from the Red Book of Hergest, and has been published by me in the Cymmrodor, Vol. iij. 52—61: it now forms an appendix to the R. B. Mabinogion, pp. 297—309.
  165. Nennius calls him Bellinus, son of Minocannus, and makes him king of Britain in the time of Caesar, with whom he fights: see San-Marte's Nennius & Gildas, pp. 40-1 (§ 19).
  166. Bk. of Leinster, 4a, and elsewhere.
  167. See folio 60b of the Bk. of the Dun, a MS. compiled, about the year 1100, from older sources: its Irish name is Lebor na huidre or Leabhar na h-Uidhre. My references are to the lithographed facsimile published by the Royal Irish Academy (Dublin, 1870).
  168. Bk. of Leinster, 254b.
  169. Battle of Magh Leana, ed. by O'Curry (Dublin, 1855), pp. 30, 36.
  170. Geoffrey's Historia Begum Britanniae (ed. San-Marte) x. 3. M. d'Arbois de Jubainville has, by a different route, arrived at the same conclusion as to the meaning to be attached to the term Spain in such contexts: see his Cycle Mythol. pp. 85, 137.
  171. This etymology was suggested to Dr. Stokes, who has approved of it and explained by means of it, in his Celtic Declension, p. 37, the irregularity of an Irish duine making dóini in the plural: the former, according to his rule, comes from an oxytone dunjó-, and the latter from a paroxytone dúnjo-. The Welsh dyn postulates the latter; but we have a trace of the former in a 'dyneᵭ,' from which was sometimes formed a plural 'dyneᵭon,' written dyneton in the Black Bk. (Skene, ij. 29, line 11), and dynedun in the Bk. of Taliessin (Skene, ij. 196, line. 10). The Celtic root being dwan or dvan, the evolution of dóini from it has its parallel in Old Irish cóic, 'five,' from a base corresponding to the Latin quinque. The Welsh dyn is now masculine, and a feminine dynes, 'woman,' has been made for it, but so recently that it is not yet a book-word; but in the Middle Ages dyn was often used in the feminine by the poets, and it occurs of that gender also in old Cornish (see Stokes's Beunans Meriasek, verse 1006): this means that the word was originally, like θνητός, an adjective, donjos, donja, donjon, of three genders. Lastly, we seem to have an element of the same origin in the Dontaurios of a magic spell extant in the Gaulish language: see Stokes's Celt. Decl. p. 78. For a different account of the origin of θάνατος, see Brugmann's Grundriss der vergleichenden Grammatik der indogermanischen Sprachen (Strassburg, 1886), i. §§ 236, 429b, 1.
  172. See Stokes, in the Rev. Celt. v. 362, note 3.
  173. Bertrand, pp. 9, 10.
  174. Bertrand, pp. 22-7.
  175. Ib. p. 23.
  176. The horns are gone, but the sockets remain, with a trace of iron rust in them.
  177. Bertrand, pp. 9, 10.
  178. Ib. pp. 24-6.
  179. R. B. Mab. pp. 34-8; Guest, iij. 103—129.
  180. The story-teller experienced some difficulty in describing them: he makes Brân's ships sail across them; then he says they were but two rivers, called Lli and Archan; and he dismisses them with the significant remark that it is since then "the Sea has multiplied his realms:" see R. B. Mab. p. 35; Guest, iij. 117, where the passage is egregiously mistranslated.
  181. See Lewis Morris' Celtic Remains, edited for the Cambrian Arch. Association by the Rev. D. Silvan Evans (London, 1878), p. 436.
  182. Skene, ii. 203-4.
  183. The Brython for 1861, p. 312a, where the name of the cave is wrongly printed as that of Maen Cymrwd instead of Maen Cymwd, for which information I am indebted to the kindness of the editor of the Brython, Mr. Silvan Evans. Unfortunately, neither he nor I can find where the cave was situated, but I have in my childhood heard similar descriptions of the devil, though I can no longer localize them.
  184. See pp. 129-30 of O'Donovan's translation (ed. Stokes), s.v. Ore Tréith.
  185. Rev. Celt. iv. 26.
  186. Ib. p. 27.
  187. Ib. p. 33.
  188. Stokes, Celt. Declension, p. 62.
  189. Rev. Celt. iv. p. 30.
  190. Ib. p. 31.
  191. Ib.
  192. Orelli, No. 2090.
  193. Rev. Celt. iv. 34.
  194. Rev. Celt. iv. 30; Rev. des Soc. savantes, 1875, ij. pp. 113-4.
  195. Rev. Celt. iv. 34.
  196. Ib.
  197. Ib. p. 35; Brambach, No. 469.
  198. Corpus Poet. Bor. i. 193, ii. 621.
  199. Hübner, No. 507.
  200. By Andrew Lang in his Custom and Myth, p. 178.