Gaelic Folklore (37): The Gods and the Supernatural

The Gods and the Supernatural 

The Nature of Celtic Religion

Because the pagan Celts did not write about themselves, the only way that modern scholars can learn anything about their belief-systems is by constructing hypotheses based upon archaeological sources and historical documents written by contemporary, but alien, classical observers, who selected and often misunderstood what they recorded.

There is another group of documents, those which tell of the earliest Welsh and Irish myths and written in the vernacular. But these have to be treated with extreme caution as sources for pagan Celtic religion, and as a strand of evidence which must be treated separately from contemporary data. This tradition is the work of Christian redactors writing in medieval times, and close links between the undoubted mythology it contains and the evidence which is synchronous with the pagan Celts (around 500 Be to AD 400) cannot usefully be made. Moreover, this vernacular tradition relates only to Wales and Ireland, far from the continental heartlands of the early Celts.

The picture painted by the evidence is of a rich and varied religious tradition. This variety and complexity is due largely to the essential animism which appears to have underpinned Celtic religion, the belief that every part of the natural world, every feature of the landscape, was numinous, possessed of a spirit. These natural forces were perceived as capable of doing humankind good or harm, and so they had to be controlled and their power harnessed by means of divination, sacrifice and other propitiatory rituals. Sacred space could take the form of built shrines, but equally important were natural cult loci such as lakes, springs and trees or open-air enclosures where worshippers were not cut off from the numinous landscape around them.

The perception of the supernatural as being present in the natural world penetrated all aspects of Celtic belief. Thus, the most popular, pan-tribal deities - the celestial gods and the mother-goddesses - were linked to their respective functions as providers of light, heat and fertility. The great healing cults of Romano-Celtic Gaul and Britain were centered upon the natural phenomena of thermal springs. Many divinities - such as Epona, Arduinna, Nodens and Cernunnos - had a close affinity with animals: indeed the horned and antlered gods took on the partial personae of bulls or stags. The perception of spirits in the landscape is amply demonstrated by the names of gods on epigraphic dedications of the Romano-Celtic period, which betray their topographical character, as personifications of places: the identity of gods such as Glanis of Glanum and Nemausus of Nemausus (Nimes), both in Provence, were merged inextricably with their locality.

Cernunnos

Problems of evidence

Sequana

 Leaving aside the difficulties of the written sources for the present, we   have a very real problem in the archaeological (that is the epigraphic   and iconographic) evidence. This results from the fact that the great   majority of inscriptions and images date to the Romano-Celtic rather   than the free Celtic (i.e. pre-Roman) period. There is considerable   overt Roman influence on indigenous religious expression. Gods in   classical guise and bearing Roman names were adopted by the Celts   but with subtle changes, such as the addition of a native surname to a   dedication, or of Celtic symbols - like the wheel or torque - to the   imagery of Jupiter or Mercury. Other divinities, like Cernunnos, the   Mothers, Epona, Sequana and Sucellus, appear almost wholly   indigenous in concept, but their imagery still owes something to   Roman iconography and, of course, their names use Latin forms, even   though their meaning may be Celtic. But even these apparently native   gods are sometimes argued to be the products of Romanization in that,   with a few exceptions, they do not appear in the archaeological record   until the Roman period.

What is perhaps more likely is that the perceptions underlying the identity of these spirits were always present but only took the form of physical images once the stimulus of the great Mediterranean traditions of iconography took hold of the Celtic world. The lack of free Celtic epigraphy and the comparative paucity of pre-Roman religious imagery means that, inevitably, we have to observe Celtic gods through a window created by Rome.

Pre-Roman iconography and cult expression.

Iconography

When an invading group of Celts overran and plundered the sacred Greek site of Delphi in the early third century BC their leader, Brennus, laughed at the anthropomorphic images of the Greek divinities which adorned the great sanctuary. He was apparently scoffing at the naivety of Mediterranean perceptions of the divine. But whilst it is undoubtedly true that religious iconography is comparatively scarce in Celtic lands before the intrusion of Greco-Roman artistic traditions, images of Celtic divinities were nonetheless present in small numbers in the last few centuries before Christ.

Stone sculpture falls into two main distributional clusters: one in central Europe; the other in the area of southern Gaul known later by the Romans as The Province (Provence), because of their early conquest of the region (late second century BC).

The central European group includes large stone male statues, dating between the sixth and third centuries Be, which are presumed to represent gods or dead heroes. One of the earliest such figures is a huge sandstone image from a late Hallstatt burial-mound at Hirschlanden near Stuttgart. He is naked but for a conical helmet, a torque, belt and dagger. His original position may have been at the top of the mound, and he may represent the dead warrior himself or perhaps a war-god, patron of the deceased. (pic. left) The image at Holzerlingen, also in Germany and of similar date, almost certainly represents a god: a rough sandstone block was crudely hewn into a human torso, featureless except for a belt at the waist. But the statue is janiform, surmounted by a dual head bearing horns, perhaps the earliest representation of a Celtic horned god. The carved head from Heidelberg comes from a statue or pillar-stone: dating to the fifth-fourth century Be, it bears a leaf crown consisting of two swelling lobes which meet above the head, and on the forehead is carved a motif which has been interpreted by some as a lotus bud, a Greek symbol of eternity. The two features of crown and lotus may signify divinity: both recur on the pillar-stone at Pfalzfeld in the Rhineland which dates to the same period, and which bears curvilinear foliate Celtic designs out of which peer four leafcrowned human heads. Further east is the third-century Be ritual site of Msecké Zehrovice near Prague, which produced a carved stone head, originally from a life-size statue: the face is that of the typical Celt as stereotyped by classical writers (and more recently by the creators of Asterix): hair en brosse, staring eyes and flowing moustache. The image bears a heavy buffer-torque, probably a symbol of divinity or, at least, high status. It is difficult to interpret the Msecké Zehrovice head as representative of anything other than a god.

The Provencal group of sculptures dates to between the fifth and second centuries Be, and their presence may be due in part to the stimulus of mimetic representation provided by the nearby Greek (Phocean) colony of Marseilles. Such sanctuaries as those at Entremont and Roquepertuse in the lower Rhone valley have produced a rich iconography, including sculptures of crosslegged male figures, some of whom wear armour, and who may be war-gods. The severed human head is a prominent iconographic theme at these shrines, and some of these 'warrior-gods' hold severed heads in their hands, as if reflective of the divine dominance over humans in life and death. Entremont produced a carved pillar of incised human heads; both this sanctuary and that at Roquepertuse had niches, filled with the human skulls of young men killed in battle, built into the structure of the temple. The sanctuary at 

Janiform head, Roquepertuse

Roquepertuse was guarded by a janiform head, perhaps that of the presiding deity, who gazed inwards and outwards from the gateway of his sacred place.  Other stone sculptures, presumed to represent divinities, come from elsewhere in the Celtic heartland: an important example is the boar-god at Euffigneix (Haute Marne), an image which dates to the second or first century Be. The carving takes the form of a roughly hewn pillar depicting a beardless god wearing a heavy torque, with a boar in low relief striding up his torso. On the side of the pillar is an immense human eye, perhaps indicative of protection or omnipotence. The boar's dorsal crest is erect, as if to reflect aggression: for the Celts, the image of the boar was an important war symbol. The Euffigneix deity may himself be a god of war, or perhaps of hunting and wild nature.

Imagery in metal - usually bronze - also bears witness to the representation of divinities in the last few centuries Be. Small figures of warrior- or hunter-gods come from St Maur-en-Chaussee (Oise), Balzars in Liechtenstein  and Dineault in Brittany, the last-mentioned a warrior-goddess with a goose-crested helmet.

Images of animals are more common than are anthropomorphic depictions: bulls, like those from Byciskila (Czechoslovakia) and Hallstatt (Austria), which date to the sixth century Be, may symbolize the sacrifice of an animal which was central to the early Celtic economy. Boar figurines may reflect war symbolism: some, like those at Hounslow (Middlesex), Gaer Fawr and Luncani in Romania, were probably helmet crests. But the bronze boars from Neuvy-en-Sullias (Loiret) are nearly life-size and almost certainly came from a shrine. The Neuvy hoard dates to the very end of the Gaulish Iron Age.

The Gundestrup Cauldron is one of the most important pieces of pre-Roman Celtic religious art. It is a large silver-gilt vessel made up of individual plates which bear a complex iconography, including depictions of divinities. The cauldron was found dismantled in a Danish peat-bog at Raevemose Qutiand), and was probably made in the second or first century Be. The circumstances of its manufacture, use and subsequent deposition have long been the subject of controversy.

Gundestrop Cauldron

It was probably made in south-east Europe by Thracian or Dacian silversmiths. Many of the iconographic themes are exotic: lions, elephants and griffons adorn the plates. However, there are a number of unequivocally Celtic motifs: the weapons and armour belong to the Celtic world, and the deities represented may be paralleled in the later Romano-Celtic imagery of Gaul and Britain. These include the antlered god, sometimes identified as 'Cernunnos' because of a first century AD Parisian monument linking this name with a similar antlered figure. On the cauldron, the god is accompanied by a ram-headed snake, an idiosyncratic cult-animal which occurs consistently with the antlered god in Romano-Celtic Gaul. This serpent appears a total of three times on the cauldron. The antlered figure appears with two torques on the vessel, and he is thus represented on a number of later Romano-Gaulish depictions. Interestingly, as early as the fourth century Be, an antlered figure with torques and horned snakes is portrayed in the rock art of Camonica valley in north Italy. Another undoubtedly Celtic divine representation on the Gundestrup vessel is the wheel-god, depicted on an outer plate. Once again, there is good evidence for this deity in Romano-Celtic imagery. The mystery of the Gundestrup Cauldron may never fully be solved. It is possible that south-east European craftsmen were commissioned to make a great cult-vessel for a Celtic clientele. The cauldron could have been looted from Gaul by Teutonic raiders, perhaps the Cimbri, and later buried by them for safety or as an offering to their own gods.

Boar god

 Vincent Megaw has pointed out that the Euffigneix stone boar-god was perhaps   carved by a craftsman who was more familiar with working in wood than in stone.   There is no doubt that images of gods were made in wood, which usually does not   survive. That they may have been common during the Iron Age is suggested by the   chance preservation of these figures in waterlogged contexts. Five pinewood images   of naked warriors with shields and detachable phalli, set in a simple boat model,   were found at Roos Carr in the Humber estuary: they may date from as early as the   seventh century Be. A wooden female image dating to the first century Be comes   from a wicker structure at Ballachulish, Argyll; and there are other stray British   figures. Oak carvings of animals, dendro-dated to 123 Be, come from a shaft at the   Viereckschanze of Fellbach-Schmiden in southern Germany.

 

The first phase of the great healing sanctuary dedicated to Sequana, goddess of the source of the Seine at Fontes Sequanae near Dijon, dates to the interface between the free Celtic and Romano-Celtic periods. This primary, first century Be, phase is represented by more than 200 wooden votives, depictions of pilgrims and the parts of their bodies requiring a cure, which were offered to the power of the sacred spring. Of similar date (first century Be to first century AD) are the more numerous wooden images at Chamalieres (Puy-de-Dome), a holy place where two springs met. We are reminded of Lucan's statement concerning a sacred grove near Marseilles, encountered by Julius Caesar's army: The images were stark, gloomy blocks of unworked timber, rotten with age, whose ghastly pallor terrified their devotees.

We are unable positively to name or identify these pre-Roman gods whose images were carved in stone, wood or bronze because this iconography is unsupported by inscriptions or documents, although their accompanying symbolism may sometimes suggest their possible function.

Ritual Behaviour

It is clear that iconography is by no means the sole evidence for religious expression in the pre-Roman period.

Sword Llyn Cerrig Bach

 Perhaps the most prominent iron age ritual   behaviour  concerns the deposition of prestigious   objects, often of a martial nature, which were   frequently buried on dry land or in watery contexts.   These implements and weapons were often   deliberately bent or broken in order ritually to 'kill'   them and thus render them appropriate as sacrifices   to the spirits of the supernatural world. Examples of   this practice are numerous: in Britain we may cite   the deposit of metalwork, including swords and   chariot fittings, in the marshy lake at Llyn Cerrig   Bach on the island of Anglesey, which spans a   period  between the second century Be and the first   century AD. Broken weapons formed a substantial element of the offerings made at later pre-Roman iron age temples, such as Hayling Island (Hampshire). This occurrence may be closely paralleled at Gournay (Oise). Deposits in Switzerland are particularly prolific:that at La Tene, on the shore of Lake Neuchatel, is well known, but the equally important deposit of weapons and chariot equipment at Tiefenau, in the oppidum of Bern-Engehalbinsel, is less familiar. This evidence of recurrent ritual behavior may be interpreted as 'conspicuous consumption', but equally it may express behaviour associated with a warrior-cult, a practice essentially similar to that described by Caesar: When they have decided to fight a battle, it is to Mars that they usually dedicate the spoils they hope to win.

Classical writers on Celtic Gods

Although Greek and Roman writers recorded their perceptions of Celtic culture and religion, especially from the first century Be, they provide little evidence for the identity or function of Celtic deities. Indeed, such commentators as Caesar, Strabo and Diodorus Siculus are more concerned with ritual practices, such as headhunting, or religious functionaries, like the druids, than with Celtic perceptions of the supernatural world. Caesar alludes to druidic lore concerning the transmigration of souls. Both Lucan and Diodorus comment on this Celtic belief in a cycle of death and rebirth.

Esus

Where mention of the gods does occur, it is heavily overlaid by Roman conflation and misinterpretation. Thus Caesar speaks as if Celtic deities are identical with those of the Roman pantheon, giving them such names as Mercury, Mars and Jupiter. Does this mean that he was unaware of their native names, that they were deliberately kept secret, or that the resemblance with his own gods was so strong that he was concerned to record this similarity? The first century AD Roman poet Lucan does allude to three gods with Celtic names who were apparently encountered by Caesar's army in southern Gaul in the first century Be. These are Esus, Taranis and Teutates, all of whom, states Lucan, demanded appeasement in the form of human sacrifice. Lucan implies that these were important Gaulish divinities, but this is not borne out by archaeological testimony. Taranis and Teutates each occur on a handful of dedications which are scattered within the Romano-Celtic world, and the name Esus appears only once, on an early first-century AD monument in Paris.  All three names are descriptive: Taranis ('Thunderer') is tied to function; Esus means 'Lord' or 'Master'; and Teutates probably refers to the divine leadership of a tribe or tuath. Esus and Teutates are therefore titles rather than names.

Sporadic references to Celtic gods appear in the literature of the Romano-Celtic period. Tertullian and Herodian allude to the cult of Apollo Belenus in Noricum and north Italy, and Ausonius alludes to sanctuaries dedicated to Belenus in Aquitaine in the fourth century AD. Grannus is referred to by Dio Cassius  who speaks of the emperor Caracalla's unsuccessful attempts to find a cure for his physical afflictions at the temples of Grannus, Aesculapius and Serapis. Reference is made to Epona's cult by a number of sources, including Apuleius and Minucius Felix. So the classical literary sources are of very little use in establishing the identities of Celtic gods and the nature of belief. Any detail concerning a Celtic pantheon must be sought from the epigraphy and iconography of the Romano-Celtic world.

The Romano-Celtic tradition

Celtic religious expression is represented during the Roman period by two main strands of material culture: epigraphy and iconography. Inscriptions give us the names of deities; imagery demonstrates how they were perceived as physical manifestations of the supernatural. One immediate problem lies in the fact that, very frequently, an epigraphic dedication to a god occurs without any accompanying image and vice versa. So it is usually impossible to marry images with their names or names with their images. Moreover, there is often a discrepancy between the ethnicity of a god as expressed by his name and by his appearance. Thus a being with a Celtic name - such as Sequana - may look very classical, whilst a god with a Roman name - Mars or Mercury, for instance - may be depicted in wholly native style.

Epigraphy

The names of indigenous Celtic gods may contain Roman and native elements or may be purely Celtic. Thus Mars and Mercury were frequently invoked with different Celtic surnames or epithets: Mars Lenus, worshipped among the Treveri, and Mars Corotiacus, invoked in Suffolk, are just two of numerous examples. Mercury was equally diverse: his titles or surnames include Cissonius and Moccus. Jupiter was surnamed Brixianus in Cisalpine Gaul and Parthinus in north-east Dalmatia, both topographical names associated with high places: these are just two of many such titles. Apollo was worshipped in Gaul mainly as a healer: Moritasgus, Grannus, Belenus and Vindonnus are among his epithets. Sometimes the Celtic name comes first: a good example is Sulis Minerva, the healer-goddess of Bath (Aquae Sulis). The pairing of Roman and Celtic god-names is confusing and difficult to interpret. Sometimes the Celtic surname is descriptive - hence Mars Rigisamus (‘Greatest King') at West Coker in Somerset, or topographical, like Apollo Grannus at Grand in the Vosges. But Sulis was clearly a goddess in her own right, equated in the Roman period with the classical Minerva.

Sucellos

Another method of epigraphic pairing concerns the linkage of male and female divinities. A pattern may be discerned here, in that, very frequently, the male deity bears a Roman or Roman and Celtic name, whilst that of the female is wholly indigenous. Examples include Mercury and Rosmerta; Apollo Grannus and Sirona; Mars Loucetius and Nemetona. Sometimes both members of the divine couple have Celtic names: such is the case with Sucellus and Nantosuelta; Luxovius and Bricta, the local spirits of Luxeuil; or Ucuetis and Bergusia, the craft deities of Alesia. Again the native element may often be interpreted as descriptive: 'Sucellus' means 'The Good Striker' (and his iconographic image is that of a man bearing a long-shafted hammer); Nantosuelta's name means 'Winding Brook'; Rosmerta is 'The Great Provider'; Nemetona 'The Goddess of the ’Sacred Grove'. Interestingly, goddesses like Rosmerta and Sirona may be invoked alone, without their partner, thus signifying their independent status within the Celtic pantheon.

Epigraphy gives us the names of many more purely native divinities, sometimes linked with images. Such is Epona, the great horse-goddess, worshipped all over the Celtic world. The mother-goddesses or deae matres are interesting: often they are known merely by their Latin title 'matres' or 'matronae', but they may bear descriptive epithets which bear witness to localized versions of their cult. The Rhineland mother-goddesses bear outlandish-sounding topographical surnames, such as the Matronae Aufaniae or the Vacallinehae.  The Celtic thunder-god Taranis is sometimes equated with the Roman Jupiter, but his occurrence alone on several inscriptions argues for his independent identity. The god of the Lydney (Glos.) sanctuary, Nodens, is invoked on his own, but he is also linked with both Mars and Silvanus, as if the native god were perceived as possessing an affinity with the functions of both Roman deities. This apparent confusion in pairing recurs, for instance, with Mars and Mercury in Gaul, whose native surnames are sometimes shared: thus both Mars Visucius and Mercury Visucius were invoked on dedications.

The Evidence of Iconography

The imagery of Celtic religion in the Roman period is rich and varied. If it is accompanied by a dedicatory inscription, a sculpture or figurine may be positively identified; if not, the symbols accompanying the image must be used to attempt some classification in terms of character or function. A depiction may be associated with an inscribed name on only one or two stones, although the image itself may appear many times. In

these instances, scholars have tended to use the inscribed name to identify similar images where the dedication is absent. Such is the case with Epona, the horse-goddess, whose image (a woman riding side-saddle on a horse or sitting between two or more horses) is far more common than are epigraphic dedications bearing her name.

Likewise, the inscribed name 'Cernunnos' accompanies a depiction of an antlered, torque-bearing god on an early first-century AD stone in Paris. But there are many images of a similarly antlered being from Romano-Celtic Gaul which bear no name. Are we justified in assuming these portrayals also represent Cernunnos? The name itself merely means 'Horned One', and so it is less a true god-name than a descriptive title. The names Sucellus and Nantosuelta occur at Sarrebourg near Metz, on a carving of a male and female, the most distinctive accompanying symbol being the longshafted hammer borne by Sucellus. But many other images of a similar divine couple were the focus of veneration in Gaul and the Rhineland, without the identifying names, although Sucellus is mentioned on one or two scattered dedications in Britain and Gaul.

The symbols or motifs which accompany images of Celtic deities give some clues as to their function or identity, although our interpretation of these symbols may be open to misconception. The celestial god frequently carries a solar wheel as his main attribute. The fact that this wheel-god is sometimes invoked under the name of Jupiter, the Roman sky-god, makes the native divinity's identification secure. The Celtic god of the sky and sun was sometimes invoked as a divine horseman, the horse symbolizing the prestige and swiftness appropriate to the high god of sun and firmament. Although this equestrian deity is entirely indigenous to the Celtic world, in terms of image and meaning, he is nevertheless venerated under the Roman name, Jupiter. The triple mother-goddesses may be represented by name, but their imagery alone betrays their responsibilities as promoters of fertility. Thus they may be portrayed with babies, children, fruit, corn or loaves. The Burgundian Mothers are often represented one with an infant at the breast, a second with a napkin, and the third with a bath-sponge and basin, and a cornucopia is frequently present in emphasis of the symbolism of abundance. Occasionally, this simple fertility imagery is subtly changed, so that the symbols of fecund plenty are accompanied or partially replaced by motifs of the Fates, such as a balance or spindle, reflective of the thread of life. In these images, the rolled napkin may instead be interpreted as a scroll, the Book of Life.

The Celto-Germanic mother-goddesses are distinctive in their imagery: they never reflect human fertility, but instead their attributes are those of fruit, bread or coins, symbolic of prosperity and commercial success. Moreover, these Rhineland goddesses may represent different ages of womanhood: the central goddess is always depicted as a young girl with flowing hair, whilst her flanking companions are older and wear huge beehive-shaped bonnets or headdresses. The triplistic character of many mother-goddess images expresses the power of 'three' in Celtic religion, a significance which transcends mere intensity of expression by means of repetition.

Triplism is an important characteristic of Celtic religious iconography, and triadism may be observed in imagery other than that of the Mothers. The Genii Cucullati were hooded dwarves associated with fertility, and indeed with the Mothers themselves, and they usually occur in threes. Gods with triple heads were depicted especially among the Remi, the Aedui and Lingones; and a triple-horned bull was venerated particularly in north-east Gaul.

The symbolism of the healer-deities is varied and interesting. The Celtic Mars, at such therapeutic shrines as Mavilly (Cote d'Or) and Trier, is not a warrior in the true, Roman, sense, but instead he fights and protects against ill-health and barrenness. Sometimes, his image is that of a soldier, as at Mavilly. Frequently, the motifs of healing and fertility are blurred and merged. Thus curative goddesses such as Sirona and Damona are represented with ears of corn, eggs and snakes: the corn is a symbol of plenty; eggs have strong fertility associations, but may also represent death and regeneration (since the egg must be broken in order to release new life). Because of their habit of skin-sloughing, serpents were clear symbols of rebirth. Many curative deities were partners, such as Apollo and Sirona: in their imagery, it is often the goddess who possesses the symbolism evocative of function.

But many healer-goddesses themselves carry no emblems which in themselves are indicative of their curative function: Sequana, the divine healer of the Seine at its spring-source near Dijon, is depicted as a woman in a long robe and a diadem, sailing in a duck-prowed boat to reflect her aquatic symbolism. But it is the presence of dedications and votive offerings that identify Sequana as a healer.

Perhaps the most powerful group of images is that associated with animals. The close relationship between god and beast is clearly reflected in iconography which displays the ubiquity and cult importance of animal symbolism. Some creatures accompanied anthropomorphic images, presumably to demonstrate a particular quality or feature associated with the god's character, just as occurs in classical imagery. But there are indications that animals in Celtic religion achieved a status denied them in the Mediterranean world. Some divinities - and Epona is a prime example - are dependent upon animals for their iconographic and epigraphic identity.

Thus, Epona is always depicted riding on a mare or accompanied by horses. Moreover, her name is philologically linked with epos, a Gaulish word for 'horse'. She was the goddess of the craft of horse-breeding, and she was revered by cavalrymen as a divine protectress of them and their animals. But she also possessed wider responsibilities as a deity of fertility and general well-being. There was even an underworld dimension to her cult. Other, less widely known goddesses enjoyed a similar affinity with beasts: Arduinna, the boar-deity of the Ardennes Forest is one; Artio, bear-goddess of Muri in Switzerland, is another.

Artio

It is difficult to establish whether or not the animals themselves possessed divine status; the likelihood is that they were sacred only inasmuch as they symbolized certain features of a particular cult. But there is debate over the status of certain creatures which appear in the iconography: monstrous animals, like the ram-horned serpent and the triple-horned bull, may well have been worshipped as beings of tremendous power, because of their hybrid or unnatural imagery. The triple-horned bull is not associated with any particular anthropomorphic god, but the snake frequently accompanies the antlered god, and is also sometimes linked with Celtic versions of Mars and Mercury. Both images occur mainly in north-east Gaul, although outliers occur as far west as Britain. Both these creatures are endowed with extra features in order to increase their potency: horns are symbols of power and fertility. Possibly linked with the triple-horned bull in some manner is Tarvostrigaranus, the 'Bull with Three Cranes', who is named and depicted on a first century AD stone from Paris.

The sanctity of animals is seen at its least equivocal in the iconography of deities whose images, although essentially anthropomorphic, nevertheless incorporate animal features. Of these, the most important are the horned or antlered gods. Images with bull- or goat-horns appear all over the Romano-Celtic world. A particular group occurs in North Britain, among the Brigantes, where local deities were depicted as naked warriors, often ithyphallic, with bull-horns. A god invoked especially in north-east Gaul is portrayed with antlers, and other recurrent features may also be discerned on these images: these include a cross-legged seating position; the possession of two torques (one worn, one carried); accompaniment by a ram-horned snake and/or stag; the possession of attributes of plenty, such as money, corn or fruit. As discussed above, a god with torques and antlers was invoked as Cernunnos on a Parisian monument, but all other images of an antlered god are without dedications. This image is particularly interesting because, unlike most Romano-Celtic iconography, there are examples which pre-date the Roman period: one of the iron age rock carvings at Val Camonica in north Italy, dating to the fourth century Be, depicts a standing figure with antlers, torques and a horned serpent; and the same god appears on the Gundestrup Cauldron, as we have seen. The presence of semi-zoomorphic images serves to emphasize the lack of rigid boundaries between animal and human which is central to early Celtic religious perceptions. Beasts were revered for their specific qualities (speed, virility, aggression or beauty) and these qualities were woven into the Celtic expression of the supernatural.

The Gods of the earliest written myths

Here, I would like to draw attention to the deities themselves who are presented in this early Celtic literature. As mentioned at the beginning of this chapter, it is important to recognize the impossibility of making direct links between the gods of these western myths and those expressed by the images and inscriptions which are the main concern of this survey. There is a wide spatial and temporal chasm between the evidence for Romano-Celtic religion and the Insular divinities to whom we are introduced in - say - the Irish Book of Invasions.

The early Welsh and Irish deities have names and characters, but no idea of cult or ritual associated with them is present in the documents. They are supernatural heroes rather than true objects of belief. It is within the Irish vernacular tradition that the pagan gods are most clearly developed. The most important are those described in the Book of Invasions, the members of the Tuatha De Danann, the divine race of Ireland, the 'people of the goddess Danu'. The Tuatha De consist of deities with specific functions and responsibilities: these include the Daghdha, who was a specialist in druid lore and magic; Dian Cecht, the physician; Goibhniu, the divine blacksmith; Lugh, the warrior and god of light, who was also skilled in arts and crafts; Brigid, a triple goddess of poetry, prophecy and fertility. The Book of Invasions recounts a series of mythical occupations of Ireland (in order to explain the presence of the Gaels or Celts). The Tuatha De were one invading group, who inhabited the island until dispossessed by the Gaels and forced to create a new magical domain underground. This Otherworld was perceived as a mirror-image of earthly life, but better, a land of immortality, joy and plenty.

The Ulster Cycle, too, had its supernatural beings, though they are portrayed as heroes rather than gods sensu stricto. Chief of the Ulster tales is the Tain Bo Cuailnge, and here we are introduced to such characters as Cu Chulainn, the archetypal champion, and the warlike and promiscuous Medb, queen of Connacht, an euhemerized goddess. These beings are essentially similar to the heroic characters of the Four Branches of the Mabinogi and the Tale of Culhwch and Olwen. The Insular and Welsh mythological traditions contain many common elements: supernatural events; magical cauldrons of regeneration and plenty; beings of superhuman size, strength and wisdom; a brilliantly portrayed Otherworld; the potency of triads; supernatural animals, and the related phenomenon of shapechanging between human and animal form.

A Celtic pantheon?

Leaving aside the separate and contentious issue of the vernacular mythology touched on above, we may legitimately pose the question as to whether it is possible to discern a hierarchy of Celtic divinities from the bewildering array of epigraphic dedications and iconographical forms of pagan Celtic Europe with which we are presented.

Jupiter column

Judging from frequency of occurrences and from distribution, it is possible to make some assessment of the popularity of various god-types. At the top are deities whose names and images appear widely and often throughout the Celtic world. These include the sky- and sun-god, although certain aspects of his cult, such as his depiction as a celestial horseman on the 'Jupiter columns' of eastern Gaul and the Rhineland, may cluster in specific regions. The mother goddesses and Epona likewise transcend tribal boundaries and appear to have been venerated by many different communities. The cults associated with these deities not only spanned wide areas but their worship percolated down from the highest to the humblest echelons of society. For example, some of the Rhineland mother-goddesses were invoked by high-ranking officials in the Roman army or civil magistrates, whilst some Gaulish and British carvings of the goddesses were clearly commissioned by groups of rural people or by a single family for veneration in private shrines. Likewise, the 'Jupiter-Giant columns' set up in honour of the Celtic sky-god were the result of corporate religious activity, but the small pipe-clay figurines of the same god from Gaulish factories would have been purchased by individuals who perhaps could not afford an altar or a bronze statuette.

Less universal than the major, pan-tribal cults were those which had specific centres of popularity but which also occur sporadically elsewhere. The healers Apollo and Sirona had an important temple at Hochscheid in the Moselle Basin and were venerated particularly among the Treveri and the neighbouring Mediomatrici.

But the couple appears also in Burgundy, at Maiain, and Sirona was worshipped as far apart as Brittany and Hungary . Another important divine couple, Mercury and Rosmerta, were prominent especially in central and eastern Gaul, but with a cluster of British monuments among the Dobunni of Gloucestershire. Images of the antlered god occur mainly in eastern Gaul, appearing, for instance, among the Burgundian tribes at such places as Beaune and Etang-sur-Arroux and among the Remi at Reims. But the Santones of Saintes in Aquitaine in western Gaul also venerated the antlered god and his image is even known among the Dobunni of western Britain, at Cirencester.

Certain Celtic divinities were the focus of cults which were of major importance but their veneration was centred upon a particular religious site of which the deity was the resident spirit. Examples such as de Sequana at Fontes Sequanae; Nodens at Lydney on the river Severn and Sulis at Bath. Lenus Mars was a Treveran god, with major sanctuaries at Trier itself, at Mohn and Pommern. But he was worshipped far away from his homeland, at Caerwent in Gwent and at Chedworth in Gloucestershire.

More localized still were the numerous obscure deities who are perhaps mentioned on only one or two dedications and who were apparently venerated solely by the inhabitants of one small settlement, or at the site of a particular spring or mountain. Such was Souconna, the spirit of the river Saone at Chalon or Fagus ('Beech Tree') in the Pyrenees.

Conclusion. 

One of the dangers of a survey such as this is the unwitting presentation of a picture which can assume a timeless continuum, spanning nearly a millennium. This is not the intention, but in part is the inevitable result of studying a period which is essentially prehistoric, where chronology is often imprecise, especially where iconography is concerned. Clearly, it is possible to make distinctions between religious behavior before and after the coming of Roman traditions to Celtic lands. It is possible to observe that imagery increased towards the end of the Iron Age, when Greco-Roman concepts and customs were already intruding upon the Celtic world. Imagery and epigraphy which come from well-excavated sites offer an opportunity for close dating. But all too often, good archaeological contexts are absent for iconography, and dating by style alone is neither easy nor reliable. It is important, however, to acknowledge pagan Celtic religion as a dynamic force, which was constantly changing and responding to the stimuli of new concepts and ideas, whilst still retaining a core of conservatism. It is indeed the tension between tradition and innovation which gives Celtic religion its essential character of diversity and enigma.

(The Celtic World, 1995 Miranda Green)

Gaelic Folklore (30): Ancient Celtic Goddesses

 Ancient Celtic Goddesses 

The extant remains of the Celtic forms of religion afford abundant testimony to the great variety of divine names which were associated therewith. No student of Celtic religion can fail to be impressed with the number of Celtic deities, who appear to have been local or tribal in character.

Even where a certain deity appears to have become non-local, it will generally be found on investigation that the sphere of the extended worship has fairly well-defined areas and centers. Some of the names of Celtic deities fortunately bear in their very forms unmistakable evidence as to their original character, so that we have an insight not only into the religious conceptions of the later and more developed stages of civilization, but also into the earlier ideas from which they sprang. It is possible, also, to see in this way how variation in the degree of civilization both locally and temporally is reflected in the forms of religion.

In dealing with the religion of the Celtic world, it is of prime importance to bear in mind at the outset that Celtic civilization was very far from being homogeneous in character, and that we must expect to see this absence of homogeneity reflected also in the religious evolution.

The inscriptional evidence that has come down to us is associated almost entirely with districts that were highly Romanized. We are thus enabled to see what districts during the Empire were those in which the conceptions of certain deities had been so developed as to make it possible to identify them with the gods of Rome. These, too, were the chief districts of trade and commerce among the Celts, and, though the earlier and the later trade-routes through Gaul, for example, were not the same, yet the routes which were important in Roman times had already risen into some prominence in the earlier period. Archaeology cannot render greater service to the study of Celtic religion than by mapping out the distribution of Celtic civilization.

The religion of the rich is never quite the same as that of the poor: that of the farmer is never absolutely identical with that of the sailor: that of the townsman always differs in some degree from that of the countryman: the merchant and the craftsman do not usually worship quite the same gods as the soldier. The gods of fashionable health-resorts vary no less than their worshippers, and the study of the forms and degrees which civilization assumed in Celtic lands may help to bring these variations more and more home to us.

Attention will be directed to the Celtic goddesses in the various districts from which evidence concerning them has come down to us. It may here be stated at the outset that there are great gaps in the available evidence concerning the goddesses in question; for example, our information from the western part of Gaul is extremely slight, while from the district of the Pyrenees, though the names of gods abound, practically no names of individual goddesses have survived. These facts will become clearer, when we take the various inscriptional zones in order, as follows:

(1) The Pyrenees,

(2) Gaul south of Lyons,

(3) Gaul between Lyons and the lower Rhine,

(4) Other districts of Transalpine Gaul,

(5) Cisalpine Gaul,

(6) Britain,

(7) Noricum and the Celtic zone of the Danube.

1.The Pyrenees district.

The only goddesses with a distinctive name known from inscriptions in this district are the Niskai of the Amelie-les-Bains tablets in the Pyrenees Orientales in the ancient territory of the Sordones. These tablets are eight in number. Some of the words, such as kantamus, rogamus, sanate, omnes, non, amiki, illius, quidquid, si, are clearly Latin, but there are other words which are apparently in some form of Celtic speech. One of these words, peisqi, raises problems similar to those raised by the Coligny Calendar and the Rom tablets. The identification of Niskas with the root contained in the English 'nixies'= water-sprites, following as it does the word kantamus, is not at all improbable. The river-name

Vernodubrum from this district is undoubtedly Celtic, but the names Iliberris and Caucoliberris seem to be of a non-Celtic form. In the territory of the Sordones, where grouped goddesses are frequent, the nearest to the Sordones being the Menmandutae or Menmandutiae of the neighbourhood of Narbo. From the Pyrenaean district west of the Sordones no distinctive names of grouped goddesses have come down to us.The Menmandutae are commemorated on an inscription at Beziers, Herault, where their name occurs in the dative on a votive tablet. A somewhat similar but unexplained name, 'Minmantiis’, has been found on an inscription at Perigueux.

2. Gaul, south of Lugdunum (Lyons).

In this district the names of goddesses are frequent. They fall naturally into two main types, those of grouped goddesses and those of individuals. The former are often termed Matres or Matrae. With this name may be compared one of the Welsh names for the fairies, ' Y Mamau ' (the mothers), a name which survives in the expression ' Bendith y Mamau,' the blessing of the mothers, used of fairy benefactions, and in that of ' Y Foel Famau,' the hill of the mothers, the highest point of the Clwydian range in Denbighshire.

We find the Eburnicae matrae at Yvours, on the Bhone, near Lyons, on the inscription restored as Matr[i]s, au[g(ustis)] Eburnici[s], etc. These appear to have been the local tutelary deities of Yvours. Recorded on inscriptions are the Mairae of the neighbourhood of Dijon (Dibio).

To the south of the Eburnicae Matrae are the Obelenses Matres at Crossillac, and the Nemetiales at Grenoble, Isere. The ' Obelenses Matrae ' appear to derive their name from a place, 'Obelum' or 'Obela,' and seem to have been local tutelary goddesses; they are called on the inscription in the dative Matris Aug(ustis) Obelesibus. The Nemetiales are also called Matres, and were doubtless the protecting goddesses of the 'nemeta ' or sacred enclosures. Their name occurs on an inscription in the dative as Matris Nemetiali[b(us).

The inscription Icotiis may represent 'Icotiae' or 'Icotii.' It occurs at Cruviers, dep. Gard. The Tangonas [Matres] are mentioned on an inscription at Venasque, dep. Vaucluse,  in the dative [Matribus] Tangonis. The name 'olatonse' occurs on an imperfect inscription at Nimes in the dative olatonis. On an inscription of Le Plan d' Au(l)ps, arrond. de Brignolles,  which reads Matribu[s] Almahabu[s], etc., we have an adjective applied to Matres which contains an intervocalic 'h' such as is characteristic of the inscriptions found in the neighborhood of German territory. In Nimes itself, a great center of the Matres worship, we have the Matres Namausicse of the Marpefio Na/xavcri/ca/So inscription. There is a specifically designated group of Matres in Southern Gaul: Matres Gerudatiae, found in the dative 'Matribus Gerudatiabus’ on a votive tablet at Saint-Esteve, dep. Var. The inscriptions the Matres in this district have no local designation. For example, at Aix one inscription has the words Matrib[us] conservatricibus, while another has Matrib[us] simply:

(1) Saint-Henry de Ventabren: Ma[t]ris,

(2) Aries: 'Mat]ribus' are associated on an inscription with Fortuna Arelatensis,

(3) Apt and at Montbrun: Matribus,

(4) Vaison: Matrabus on three inscriptions, Matribus on three, and Matris on one,

(5) Sahune and at Dieule-fit,  dp. Drome: Matris,

(6) Vienne, Lyons, and Grenoble: Matris augustis,

(7) llon-daz: Matris Mithres,

(8) Saint-Innocent: Matris au[gustis],

(9) Geneva: Matr(ibus) aug(ustis),

(10) Brienne: Mat(ribus),

(11) Nimes: Matris,

(12) Narbonne: [M]a[t]ribus,

(13) Lyons: M. a. t. r on and Matris and Matribus, 

(14) Moirans, Saint- Vit and Besancon: Matrabus,

Along with these examples may be given that of Le Bourget, where we have a dedication 'Mercurio et Matr[is],' and that of Allmendigen, near Thun, where we have Iovi, Matribus, Matronis, Mercurio, Minervse, Neptuni, and once more Matribus and Minervse. It is clear from this that Lyons and the districts around it were prominent centres of the Matres worship.

The three inscriptions of Spain at Porcuna, where M alone represents Matribus, which is qualified by Veteribus, that of Duraton (Matribus) and that of Muro de Agreda, prov. Soria, where we have Matrubos, are isolated and stand on a somewhat different footing from the others.

The grouped goddesses of the south of Gaul are not, however, Matres only. We have here also in great prominence, especially in Nimes, the Proxumse or Proximse,

(1) Saliers, Bouches-du-Rhône, at Vaison,  Nimes and Clansayes: Proxsumis suls,

(2) Avignon: Proxsumis,

(3) Lourmarin: P(roxumis),

(4) Orange: Proxs(umis),

(5) Vaison: Proxsumis and Proxumis,

(6) Beaucaire: Nimes: Proxum(is),

(7) Barron: Proximis Ledaa,

(8) Nimes: [Pr]oxxumis, Proxsumis, in various degrees of completeness, and Proxumis, more or less completely.

With the Matres, as with the Matronoe of Cisalpine Gaul, may be associated the grouped goddesses called Junones, who are mentioned on a votive inscription at Nimes, as Junonib(us) Montan(is), and at Aigues-Mortes, Gard, as Junonibus Aug(ustis). These Junones appear to have been also worshipped in the zone of the central health resorts; for example, an inscription at Neris-les-bains, dep. Allier, of the Antonine period, reads 'Numinibus Augustorum et Iunonibus Neriomagienses.' On an inscription also between Langres and Toul we have Deabus Iunonibus, at Bordeaux  there is an inscription Iunonibus Iulise et Sextiliae.

In Southern Gaul, as in Spain, we also find the worship of the nymphs prominent under their Latin name:

(1) Saint Saturnin-d'Apt, Lez, Castel: two inscriptions each,

(2) Vaison: two, one being dedicated Nymphis Augustis and also an inscription with the same words at Bourguin,

(3) Goul, Apt, Carpentras, Rasteau, Vercoiran, Castera-Veveril, Castillon, Alzey and Mombach: one each,

(4) Les Fumades, near Allegre, dep. Gard: six inscriptions, one of which has the words Nymphis Augustis,

(5) Uzes: there is a record of the dedication of a temple to the nymphs in the first century a.d,

(6) Nimes: five votive tablets to nymphs occur,

(7) Puech, in Provence, there are three votive tablets to nymphs.

In the zone of the south of France, too, we find another name of a group of goddesses like those already described, in the case of the Baginae, the companions of Baginus, the local deities of Mt. Vanige and the village of Besignan: the names 'Bagino et Baginahabus' occur on an inscription at Bellecombe, canton du Buis.

Dep.Gard: There is a goddess Perta commemorated: the name occurs in the dative Pertse.

Aix-en-Diois, Drome: Bormana, the companion of the god Bormanus occurs on an inscription 'Bormano et Bormana',

Saint Vulbaz (earlier Saint Bourbaz), canton Lagnieu: Bormana on an inscription of the phrase Bormanae Augustas.

Saint Saturnin d'Apt: an inscription to another goddess, Albiorica, whose name is given in the dative Albiorice. She was doubtless regarded as the companion of Albiorix.

Viens, dep. Vaucluse,  there is an inscription to Bergonia, a name doubtless cognate with Brigindu and Brigantia.

Pertuis, dep. Vaucluse, arrond. Apt (Dexsivse), and at Cadenet in the same district, on the hill of Castelar, three times: The goddess Dexsiva, she may either be the eponymous goddess of the Dexivates, or a goddess of fortune.

Trets: Trittia, the local goddess of an old town in the arr.of Aix,

Carnoules, dep. Var: Trittiae,

Pierrefeu, dep. Var Trittae.

Montaren near Alais, dep. Gard, Saint Honore-les-bains, dep. Nievre: [R]iton[a]e, the name Ritona is the name of a river-goddess, that of the modern river Rieu.

Beziers: an inscription in honour of a goddess, who may be either Bicoria or Tricoria;

Bagnol-sur-Ceze, dep. Gard: an inscription to a goddess Diiona (cf. Dibona, Divona) This was probably the eponymous nymph of the brook la Vionne or l'Andiole, which flows into the Ceze. The name occurs in the nominative, Diiona. The form of the name suggests that in this dialect of Celtic there was a tendency to drop intervocalic ' v.'

Le Prugnon near Antibes: The remaining name of an individual goddess from this district is Thucolis. She appears to have had a priestess or priestesses.

It will be readily seen from the evidence of these inscriptions that, in the south of Gaul, group-goddesses were far more widely worshipped than individual goddesses. The latter appear to be in this district highly localized in character, and perhaps were often regarded only as members of groups. On the other hand, individual gods are numerous, while grouped gods are rare. Here, too, we do not find traces of the worship of one deity at the expense of others, but the ancient type of worship of groups seems to hold its own.

3. The district between Lugdunum and the Lower Rhine.

 This is a large district, which may be roughly regarded as extending from Lugdunum to Treves (Augusta Treverorum) and thence to Cologne and to the mouth of the Rhine. From this large and important district many inscriptions containing names of Celtic deities have come down to us, and in this respect it stands in very marked contrast to the districts in the west of Gaul. Except in the neighborhood of Cologne, grouped goddesses are here conspicuous by their absence. In their stead we find several individual goddesses, often associated with some god, and some of these goddesses appear to have attained to more than a strictly local worship.

In accordance with the plan already adopted of dealing with the group-goddesses first, the worship of Matres may be first considered.

(1) Langres: Matra[b(us), Matra, Majtris, and Ma]trab[us. These, together with the Mairse of Dijon, may be regarded as forming the northerly representatives of the group-goddesses of the Rhone valley and its neighborhood, If we now cross into the Rhine valley, we reach another zone of Matres or Matra. In this zone the Matres are generally qualified by some local or descriptive adjective.

(2) Bonn: 'Matiribus domes ticis',

(3) Andernach: 'Matribus suis',

(4) On the road from Zahlbach to Mainz: 'Iovi optimo maximo et Matribus',

(5) Frankfort on the Main: Matribus,

(6) Heddernheim: Matribus,

(7) Ell, in Alsace: Matrabus,

(8) Berkum: Matribus suis.

The Matres that are qualified by local or other adjectives are distributed as follows:

(1)Colonia Trajana (Xanthen): Brittae Matres, Maxacae matres, and Arsacse matres,

(2) Beeck, near Xanthen: Matribus Brittis,

(3) Cologne: the Malvisae, a group of goddesses commemorated in the terms 'iabus Malvisis et Silvano', (The name as ' Malvisis ' occurs also at Nieukerk Netherlands.) the Matres Mediotautehae, the Axsinginehae Matronae,

(4) Castel, near Mainz: The Ollogabiae, the name Ollogabiae apparently means ' All-seizing,' from Olios = Welsh Oil, and gab- = Welsh gaf- (in gaf-ael), and Old Irish gabim.

Another important name of group-goddesses in the Rhine district is that of Matronae, a name widely used among certain German tribes, who are thought to have adopted the religion of the Gauls. This name was also a favourite one in Cisalpine Gaul.

(1) Tetz, near Jülich: the Matronae Cantrusteihiae, (also found at Hoeylaert, near Brussels, Belgium),

(2) Altenberg, near Cologne, Rodingen, near Jülich; the Gesahenae Matronae,

(3) Rodingen and Bettenhofen: Gesahenae Matronae, the Etrahenae,

(4) Odenhausen, near Berkum: The Ascricinehae Matronae,

(5) Odendorf, near Euskirchen: The Ascricinehae,

(6) Bürgel, near Sollingen: the Matronae Rumanehae et Maviaitinehae,

(7) Sinzenich: Matronis Tummaestis.

The prominence of the worship of group-goddesses in this district is very remarkable. Of the other types of group- goddesses the Mairae of the neighbourhood of Dijon are of interest:

(1) Dijon: Dis Mairis,

(2) Til-Chatel, dep. Cote d'Or, Deae Eponae et dis Mairabus, genio loci and deabus Mairis.

In this zone there is no reference to the Proximae, but between Langres and Toul there is an inscription 'Deabus Iunonibus'.

Coming now to the Rhine valley,

(1) Cologne: Iunonibus [G]abiabus,

(2) Bonn: [Matribus or Iunonibus do]mesticis [Lugo]vibus comfedonibus,

(3) Zulpich: Iunonibus domesticis,

(4) Piitzdorf: Iunonibus,

(5) Altenhoven in the neighbourhood of Aachen: Iunonibus.

 It is not improbable that the Iunones were the Matres or Matronae of the district. On the other hand, inscriptions to nymphs are conspicuous by their absence from the Rhine valley though they are frequent elsewhere.

When we come to the individual goddesses of the zone now under consideration, we find several names which are attested by inscriptions. For example, there is Damona, the companion of Borvo, ' the Boiler,' the god of certain hot springs. The form of the name Damona suggests that it is a parallel to Epona, the former being a goddess of cattle, the latter of horses. It is possible that originally both were deities of corresponding animal form. Damona appears to be associated with the root dam-, which we find in the Irish dam (an ox), and in the Welsh dafad (a sheep), for an older damat-. The inscriptions on which Damona's name occurs are  chiefly from Bourbonne-les-Bains, dep. Haute-Marne:

‘Borvoni et Damonae,'

'Deo Apollini Borvoni et Damonae,'

'Borvoni et Damonse,'

'Deo Borvoni et Damone,'

'Borvoni et Damonae,'

‘Damonae Augustae,'

'Deo Borvoni et Damonae,’

'Borvoni et Damonae.'

The other district in which inscriptions to her are found is Bourbon-Lancy, dep. Saone-et-Loire,

'Borvoni et Damonaa,’

'Bormoni et Damonae.'

The form Bormoni shows the influence of Bormanus. Aquae Bormonis appears to have been the old name of Bourbon-Lancy. In Damona we not improbably have an old tribal animal goddess, who came to be associated with Borvo, a god of hot springs, as his name implies.

Another individual goddess, whose name appears to be associated with the zone now under consideration, is Litavis. This name occurs in conjunction with that of the god Cicolluis, identified with Mars. Some of the inscriptions on which the name is thought to occur are very imperfect:

(1) Matain, dioc. Langres: 'Marti Cicollui et Litavi, Marti Cicollui et Bellonae (Litavis was regarded as a goddess of war. 'Litavis' would appear to be identical with that of the Welsh Llydaw, e.g. a lake called Llyn Llydaw in the Snowdon district. It appears, too, to be the basis of the name of the Aeduan Litavicus or Litaviccus (Caes. B. G. vn. xxxvii.  and also of Convictolitavis, who is mentioned in the same passage. The name Litaviccus occurs also in the genitive Litavicc(i) on an inscription in the first century from Monthureux-sur-Saone, dep. Vosges. It occurs also in the dative on an inscription at Langres, as well as on silver coins of the Aeduans. A place-name of the same basis, too, Litavicrarus, occurs on an inscription at Langres in the will of a member of the tribe of the Lingones of the first century a.d. The inscription reads 'Ante ce[l]lam memoria que est Litavicrari.' The name Litavis is also probably contained in that of Cobledulitavus, found on an altar in the museum of Perigueux, Dordogne, set up by a priest of the altar of Lyons. The inscription reads [Tutelae Vesonnse] et deo Apollini Cobledulitavo . . . v. s. 1. m. The name of Litavis may be safely regarded as that of one of the most prominent deities of the Lingones, and, as derivatives of this name are found among the Aeduans, her worship probably extended to the latter tribe as well),

(2) Aignay-le-Duc, dep. Cote-d'Or: 'Deo Marti Cicollui et Litavi’.

In the zone now under consideration there are two other names of goddesses that are found mainly along with the Latin names Mercurius and Apollo. There are Rosmerta, named along with Mercury, and Sirona, named along with Apollo:

(1) Aix: Vssiae Ros[mertae?] Mercuri[o] v.s.l.m.,

(2) Gissey-le-Vieil, dep. Cote-d'Or: Aug(usto) sa[c(rum)] Deae Rosm[er]tae,

(3) Alise-Sainte-Reine: it is doubtful whether the inscription should read [Ro]sme[rtae] or Sme[rtullo],

(4) Langres: Deo Mercurio et Rosmerte,

(5) Grand: Mercurio et [Ros]mertae,

(6) Worms: Deo Mercuri(o) et Rosmerte,

(7) lzey: [Deo Merc]urio et R[osmerte],

(8) Spechbach, near Lobenfeld: [Mercu]rio [et Ros]mert(a)e,

(9) Cologne: Mercu[rio et Rosjmerte,

(10) Andernach: Merc[urio et] Rosmertae, and Me[rcurio et R]o[smertae],

(11) Huttigweiler: [Mercujrio [et Ro]sm[e]r[tae],

(12) Nider-Emmel, Zumeth, Bernkastel: D[eo] Me[rc]urio [et] Ro[s]me[rtae],  Mer[curio e]t Rosm[ertae], and Deo Mercurio et d(e)?e [R]osmertae,

(13) Reinsport on the Moselle:  Deo Mer[c]urio et Rosme[r]te,

(14 )Soulosse: D(eo) M(ercurio) et Rosmerte, and also Mercurio (et) Rosmert(ae) sacr.vicani Solimariac(enses),

(15) Mt. Sion, dep. Meurthe- et-Moselle: 'Deo Mercurio et Rosmertae',

(16) Metz: 'Deo Mercurio et Rosmertae’,

(17) Wasserbillig, Luxembourg: Deo Mercurio et deae Rosmertae,

(18) Chatenoy, dep. Vosges: ' Mercurio et Rosmertae sacrum'.

From this we may gather that Rosmerta was worshipped largely in the neighbourhood of the Rhine, and in the neighbourhood of Langres. The origin of the name is doubtful. The root may be smer(t)-, brilliant, so that Ro-smerta would mean the exceedingly brilliant one. The same root appears to occur in other proper names, such as Smertus, Smertullos, Smertomara, Smertorix, Smertuccus, Smertulitanos, and the old British tribe of the Smertae, as well as Atesmerius. We have also Smerius, Smertalus, Smertu, and Cantesmerta. From this it will be seen that the root in question was extensively distributed in personal names.

The next goddess, Sirona, is widely associated in the area in question with Apollo, her name being also written as Dirona or Dirona. The distribution of her inscriptions is as follows:

(1) Baumburg, Apollini Granno [et Si]ronae,

(2) Rome, Apollini Granno et sanctae Sironae sacrum,

(3) Bordeaux, Sironae,

(4) Luxeuil, Apollini et Sironae,

(5) Bitburg, Apollin[i Granno] et Siro[nae],

(6) Nierstein, Deo Apollini et Sironae,

(7) Mainz [Deae] Sirona ...,

(8) Grossbottwar am Marbach, a.d. 201, Apo[lli]ni et Sironae ,

(9) Maximiliansau, Deae Sironae,

(10) Wiesbaden, Sironae,

(11) Andernach DirfonaeJ,

(12) Graux, dep. Vosges, Apollini et Sironae,

(13) Sept-Fontaines, near Saint- Avoid, in Lorraine, Deae Dironae,

(14) Corseul, dep. Cotes-du-Nord, arrond. Dinan, cant. Plancoet, Sirona.

From these inscriptions Sirona may be regarded as the companion of Grannus, whose name we have in Aquae Granni (Aix-la-Chapelle) and in Granheim. The name Sirona was not improbably that of the Earth, regarded as a goddess, and it probably meant the 'long-lived one.' Another name for Apollo besides Grannos in this district is Mogounos, (whence Moguntiacum).

Of the other names of goddesses in the zone now under consideration Icovellauna  occurs near Divodurum (Metz).  Icovellauna occurs like Mogontia at Le Sablon, near Metz, and may have meant the 'protectress of health.'

Epona  is one of the most widely distributed of the names of Celtic goddesses. The goddess Epona is commemorated on numerous inscriptions :

(1) At Guidizzolo between Mantua and Verona,

(2) At Siguenza,

(3) At Also-Ilosva in Dacia,

(4) At Waitzen in Pannonia,

(5) In Carinthia,

(6) In Zollfeld,

(7) At Mariasaal, Herculi et Eponae aug(ustae),

(8) At Cilli, Eponae aug(ustae) sacrum : Iovi o(ptimo) m(aximo) Eponae et Celeise sanctee,

(9) At Windenau near Marburg in Steier-mark, Eponae aug(ustae) sacrum,

(10) At Pforing near Ingolstadt in Rhaetia, Campestribus et Eponae,

(11) At Mount Eudrik in Moesia, Epone,

(12) At Karlsburg in Dacia, Epone regin(ae) sanc(tae),

(13) At Varhely in Dacia, not long after a.d. 107, Eponabus et Campestribus,

(14) At Salona, [Iovi optim]o maxsi[mo . . . Epo]ne [. . . Marti] Cam[ulo],

(15) At Rome, associated on various inscriptions with the Matres Sulevae,

(16) At Carvoran,

(17) At Auch- indavy, near Kirkintulloch in Scotland,

(18) At Lyons,

(19) At Naix in the department of Meuse,

(20) At Metz,

(21) At Solothurn,

(22) At Til-Chatel in the diocese of Langres, Deae Eponae et dis Mairabus,

(23) At Andernach, Eponae sacr(um),

(24) At Heinzerath, Kreis Bernkastel, on two inscriptions.

There is a place called Epona, now Ep6ne, in the department of Seine-et-Oise, arrond. Mantes, and there are also two places called Eponiacum:

(1) Eppenich, near Aachen,

(2) The modern Appoigny, dep. Yonne.

Eponicus occurs as a man's name on an inscription at Rome. The zone of distribution of the Epona inscriptions will give a fair conception of the districts where her worship was most prominent. As her name shows, she was pre-eminently a goddess of horses.

Among the other names of goddesses which occur in this zone or its neighboring districts, may be instanced that of Clutoissa or Clutoidda, which occurs on an inscription not far from Noviodunum in Gaul. Her name occurs on two inscriptions, one from the village of Mesves-sur-Loire in the depart-ment of Nievre, near a spring with a reputation for healing fevers. The inscription reads, Aug(usto) sacrum, deae Cluto[i]dae, etc. It occurs, too, on a patera from Etang-sur-Arroux, dep. Saone-et-Loire. There is no evidence that Clutoida was more than a purely local goddess of a healing spring.

At Bourges there is an inscription to Solimara, a word which probably means 'the large-eyed', Solimarae sacrum. Solimaros occurs over a wide area, for example at Cilli, Sziszek, Scherschell, Martigues, Orange, Brignon, dep. Gard, near Ledignan, Nimes, Bordeaux, Paris, Breitenbach, Gustavsburg (Mainz), Heddernheim, and on the gold coins of the Bituriges Cubi (before B.C. 58), which were found at Amboise, Vendeuil Caply, dep. Oise, as well as in Yivonne (dep. Vienne), and in Vernon. A Gallo-Roman name Solimarius occurs at Apt, Bordeaux, and at Niersbach in the Prussian Rhine-province. A name of the form Solimario occurs on an inscription at Nimes. There was a place called Solimariaca on the Roman road from Metz to Langres between Neufchateau (Vosges), and Toul (Meurthe-et-Moselle).This word, too, appears to underlie the names Sommere, dep. Saone-et-Loire, Saumeray, dep. Eure-et-Loire; Saumery (Solimariaca), dep. Loiret; Le Saulmery, dep. Loiret; Saumery, dep. Loir-et-Cher; as well as the Italian Sumirago in the province of Milan in the district of Gallarate.

On the Celtic inscriptions of Volnay we have the name of a goddess Brigindu in the dative Brigindoni. It is her name that probably underlies that of Brigendonis, now Brognon, in Cote d'Or, arrond. Dijon, as well as the Ager Briendonensis in the Macon country. It is possible, too, that Gregory of Tours meant this goddess, when he said that the Gauls worshipped Berecynthia. He says that there was an image of this goddess which was carried on a vehicle to ensure the success of the fields and vineyards (pro salvatione agrorum et vinearum). Before this image the people danced and sang.

Other names of isolated goddesses are Abnoba or Deana Abnoba, the presiding deity of the Black Forest on the German side of the Rhine;

Brixia, of Brixia (Breuchin), near Luxeuil,

Aventia in the territory of the Helvetii from whom Aventicum derives its name,

Naria Nousantia mentioned on an inscription at Grissach, near Landeron, canton Neuenburg. The inscription is on a votive tablet set up by a certain T. Frontinus Hibernus,

Saint-Marcel-lez- Chalon, dep. Saone-et-Loire: a local goddess called Temusio,

Lorraine:  Nantosvelta, who is named along with a god Sucellus.

4.The Remainder of Transalpine Gaul.

In dealing with the west and north-west of Gaul we cannot but be struck by the great scarcity of names of goddesses from these districts. This is due, we may be sure, not to any absence of goddesses but to the slight extent to which the Gallo-Roman fashion of setting up votive tablets had penetrated to these regions. It is to be noted that in the neighborhood of Cherbourg the goddesses of cross-roads (Quadrivise) were objects of worship, but we have no evidence of the worship of Matres or Matronse in these districts. There is a solitary inscription to a spring-goddess Acionna at Fleuri, near Orleans, where we have the words Aug(ustae) Acionnae sacrum.

At Perigueux a goddess Stanna, perhaps a spring- goddess of the Petrucorii, is mentioned on three inscriptions in conjunction with a god Telo, the spring-god of Tolon, now Le Toulon, near Perigueux, dep. Dordogne. The root of Stanna is not improbably sta-, to stand, and may have been originally given to the earth-goddess as 'the abiding one.' The name Telo may possibly underlie the name of Toulon-sur-Mer (Telo Martius), and the place-name Telonnum, a town of the Aeduans, Toulon-sur-Arroux, near Autun, dep. Saone-et- Loire, and also the present commune called Lipostey, dep. Landes.

Another goddess of the southern area of Gaul, whose distinctive name was generally omitted, was Divona or Devona, a name which means simply ' the goddess.' The name occurs as that of the spring 'Fontaine des Chartreux ' in Cahors, dep. Lot, and then as that of Cahors itself. Divona is also given by Ausonius as the name of the spring of Bordeaux, and is probably to be read on an inscription at that town: Divionae. This form, too, seems to underlie that of the modern Divonne, dep. Ain, as well as that of Dewangen in Germania magna, given by Ptolemy n, xi, as Arjovova.

There is also a local goddess Dunisia, whose name occurs on an inscription at Bussy-Albieu, dep. Loire, The name probably means,'the goddess of fortresses.'

The name ' athubodvae ' occurs on an inscription of Fins-de- Ley, dep. Haute- Savoie. It has been thought to stand for Cathu-bodvae, and so to be the equivalent of the Irish Bodb-Catha, a goddess of war, but this identification is very far from being certain.

5. Cisalpine Gaul.

In this district inscriptions to the grouped goddesses called Matronse are numerous, for example:

(1) Verona, Iunonib(us), Matronis,

(2) Marzana, Matrona[b(us); Isorella, Matronis ; Calvisano, Matronabu[s]; Manerbio, Matronab(us),

(3) Nuvolento, in the province of Brescia, and on about forty-seven other inscriptions.

As already mentioned, the Matronae appear to have been worshipped, too, on German territory. In this district, also, the Iunones were widely worshipped, and we find inscriptions naming them about twenty-seven times.

Another group of goddesses is called Dervones or Dervonnse, ‘the spirits of the oak': these are called on the inscription of Cavalzesio, near Brescia, Fatis Dervonibus, and, on another we read Matronis Dervonnis.

With these may be compared the Silvanae, who are mentioned once on an inscription of Verona in the dative, Silvana- bus. In Cisalpine Gaul inscriptions to the god Silvanus are frequent.

Of individual goddesses Epona seems to have been worshipped in Cisalpine Gaul, as we see from an inscription at Guidizzolo between Mantua and Verona, as well as from an inscription at Siguenza. The goddess Epona is mentioned by Iuvenal, viii. 154-157, and, according to the scholiast on the passage, she was a patron of muleteers as well as of horsemen. Plutarch, Parallel, c. xxix., p. 312 E, states that she was born of a mare. Reverting to the grouped goddesses, it may be here stated that inscriptions to nymphs occur only about four or five times in Cisalpine Gaul.

6. Britain.

In Britain the grouped goddesses most widely worshipped were the Matres:

(1) Winchester: Matrib(us) Ital[i]s, Germanis, Gal[lis], Brit(annis),

(2) London: Matr[ibus],

(3) Chester, the singular Deae Matri,

(4) Doncaster, Matribus,

(5) Ribchester: Deis Matribus,

(6) Micklegate: Mat(ribus) Af(ricanis), Ita(licis), Ga(llicis),

(7) Carrawburgh: Matribus com[munibus],

(8) Aldborough: I(ovi) o(ptimo) m(aximo) Matribus M . . .,

(9)Lowther: Deabus Matribus tramarin(is), those who set it up being a vex(illatio) Germa[norum],

(10) Lough, near Plumpton Wall in Cumberland: same as 9, Deabus Matribus tramarinis,

(11) Old Carlisle: [Dea]bus Ma[tribus],

(12) Skinburness, near Silloth: Matribus Par(cis),

(13)Binchester: Mat(ribus) sac(rum),

(14)Newcastle-on-Tyne: Dea[bus] Matribus tramarinis patri(i)s,

(15) Matfen Hall: we have Deabus Matribu[s],

(16)Chesters: Deabus Matribus communibus,

(17) Housesteads:Ma[tribus] on two inscriptions, the second of which was set up by a cohort of Tungri, (18) Carvoran: two inscriptions, one reading Matri . . . , the other Matrib(us),

(19) Cambeckfort in Cumberland: we find the formula M[at]ribus omnium gentium,

(20) Walton-House-Station: Matribus t[ra]ma[rinis],

(21) Stanwix and Dykesfield: Matribu[s djomesticis, that is the guardians of the foreign soldier's home,

(22) Carlisle: Matrib(us) Parc(is),

(23) Bowness and York: Matribus suis,

(24) Bisingham: Matribus tramarinis,

(25) Burnfoot Hall and Castlecary: Matribus,

(26) Newcastle, Backworth, Matr(ibus) and Matrum.

In Britain, too, there are inscriptions to nymphs:

(1) Great Boughton: Nymphis et Fontibus,

(2) Blenkinsop Castle: Deabus Nymphis,

(3) Risingham: Nymphis venerandis,

(4) Nether Croyfarm, near Croyhill: Nymphis,

(5)Greta Bridge: the singular Deae Nympha[e],

(6) Newtown of Irthington: Deae Nymphae Brig(antia).

Whatever doubt there may be as to the local connections of some of the other deities we have here undoubtedly a goddess of Britain. It may be noted also that an inscription at Benwell, near Newcastle-on-Tyne, is set up to the three Lamias (Lamiis tribus).

Coming now to the individual goddesses, we have the following mentioned on inscriptions in Britain. Ancasta is mentioned in the formula Deae Ancastae on an inscription at Bittern, near Southampton.

The name Belisama does not occur in Britain as the name of a goddess, but only as that of an estuary, probably the Mersey or the Kibble, called by Ptolemy II. iii. 2. In the south of Gaul, however, the name occurs as that of a goddess, as, for example, on the Celtic inscription in Greek characters of Vaison, Vaucluse, ‘Segomarus, son of Villonus, a citizen of Nemausus, made for Belisama this temple. At Pont de Saint-Liziers in Les Couserans there is an inscription with the formula Minervae Belisamae sacrum. It is from her name that the place-name Belismius has probably arisen, a name surviving in Blismes (Nievre), Blesmes (Marne), and Blesmes (Aisne). The name Belisama is probably a superlative from the root bel-, which is found in the Welsh rhy-fel, war. The name Belismius occurs as a man's name on an inscription at Caerleon-on-Usk.

Of British goddesses, Brigantia is one of the most important:

(1) Greetland: D(eae) Vict(oriae) Brig(antiae) et num(inibus) A(u)g(ustorum),

(2) Adel near Leeds: Deae Brigan[tiae],

(3) Cumberland: she is called a nymph in the words Deae Nymphae Brig(antiae),

(4) Birrens near Middleby: Brigantiae s(acrum).

It is not improbable that Brigantia was the tribal deity of the powerful tribe of the Brigantes of the north of England.

Another goddess whose name occurs on inscriptions in Britain is Epona. She is mentioned on an inscription at Carvoran in the formula Deae Eponae, and at Auchindavy, near Kirkintilloch, in Scotland,

in the middle of the second century, along with Mars, Minerva, the goddesses of the fields (Campestres), Hercules, Epona, and Victory. With this inscription may be compared the inscriptions of Rome, on which Epona is mentioned along with several other deities.

A goddess Lata, or Latis, is mentioned on inscriptions at Kirkbampton: Deae Lati and Birdoswald: Dae Lati.

On two inscriptions, one at Chesters: Dea(e) Rat . . . and the other at Birdoswald: Dae Rat, a goddess is mentioned, whose name is given in an abbreviated form as Rat. There is not enough evidence to associate her with Ratae, the old name of Leicester.

It may be here noted that, in the case of the names of gods found in Britain, the chief links with the continent are for the most part with the district around Cologne, Treves, and Mayence, but there are sometimes unexpected links with other districts. Grannus ( = Apollo), who was largely worshipped in the neighbourhood of the Rhine, is mentioned on an inscription found at Musselburgh, near Edinburgh, but we find Ialonus, whose name occurs on an inscription at Nlmes, mentioned also on an inscription at Lancaster.

Leucetios, the Mars of the neighborhood of Mainz, is mentioned on an inscription at Bath.

Maponos ( = Welsh Mabon) is mentioned three times as a god in Britain, as well as in a place-name Maponi, but on the continent there are no traces of him as a god.

British references:

(1) Ribchester, county Durham: to Deo sancto Apollini Mapon(o),

(2) Ainstable, near Armthwaite, Cumberland: Deo Mapono,

(3) Hexham, in Northumberland: Apollini Mapono.

Mogons ( = Apollo) was worshipped in Britain by Vangiones in the Roman army, as may be seen from the formula Deo Mogonti, which occurs on inscriptions at Plumpton Wall (Old Penrith), Netherby, and Risingham.

The god Silvanus, too, who was widely worshipped in Spain, in Cisalpine Gaul, in the neighborhood of the Rhine, in the south of Gaul, and in central Europe, is also mentioned in Britain on about sixteen inscriptions. On one at Housesteads he is identified with Cocidius by a certain prefect of a cohort of Tungrians. This wide worship of Silvanus in the Celtic world is very suggestive of the form into which the early tree-worship of the Celts had developed.

The name Maponos, identical as it is with the Welsh Mabon, suggests that Modron, who is represented in Welsh legend as his mother, was an ancient British goddess, whose name in Roman times would be Matrona, a derivative of the root matr- (mother). This very name, it may be noted, is the original of that of the river Marne. Some of the river-names of Wales appear to be formations of this type, and the suggestion naturally arises that they also were names of goddesses.

For example, the name Aeron, in Cardiganshire, may stand for Agrona, the goddess of war.

Tarannon may have meant the goddess of thunder.

The river Dee, Deva (Welsh Dyfr-dwy), means simply the goddess.

The two streams, Dwyfor and Dwyfach, near Criccieth in Carnarvonshire, probably mean 'the great goddess' and 'the little goddess' respectively.

The name Ieithon, a stream in Radnorshire, may mean ' the goddess of speech.

'Crawnon,' in Breconshire, may mean ' the goddess of storage.'

7. The Transrhenane and Danubian districts.

In these wide zones the task of separating Celtic and Germanic deities is well-nigh impossible, but its very difficulty suggests that to both peoples the popular substratum of religion had far more in common than is usually supposed. For example, we have among the Germans as among some sections of the Celts a most remarkable development of the worship of Matres and Matronae, a form of worship of a very primitive character. This similarity of underlying religious belief is also confirmed by the study of folk-lore, as any reader of Dr. Frazer's Golden Bough can readily ascertain. At Varhely in Dacia, on an inscription made not long after 107 A.D., there is mentioned even a group of Eponae in the formula Eponab(us) et Campestrib(us) sacrum.

In Noricum again there is a group called Alaunas, who were worshipped along with Bedaios.

It may be safely conjectured that in the countries east of the Rhine grouped goddesses abounded.

Of the names of individual Celtic goddesses worshipped in these territories, especially in Pannonia and Dacia, that of Epona is by far the most prevalent.

In Noricum we find the worship of Adsalluta closely associated with that of Savus, the river Save. At Saudorfel her name occurs on five inscriptions, on two of which it is associated with that of Savus. At Hrastnigg, too, her name occurs on an inscription which reads Adsal(l)ute Aug(uste).

Another goddess who deserves mention here is Noreia:

(1) Mount Avala, near Belgrade: an inscription of the year 287 a.d., D(ea)e Nor[e]ia[e] sacrum,

(2) Hohenstein, near Pulst: Noreiae Aug(ustae) sacr(um),

(3) At Hohenstein and at Ulrichsberg: she is identified with Isis “Isidi Norei(ae)”,

(4) Trojana: Noreie August(ae),

(5) Cilli: one inscription, along with Jupiter and with Celeia, as 'Noreia sancta.' On another inscription she is mentioned along with Mars, Hercules, Victoria, and Noreia,

(6) Kerschbach: [Marti A]ug(usto) e[t NJoreiae Re[g(ina)e et] Britania[e pr]o vic(toria) L. Sep[timii Severi p]ert(inacis) inv(icti),

(7) Weihmorting, in the district of Griesbach in Lower Bavaria: Noreiae Aug(ustae) sacrum, while at

(8) Khamisa: she is mentioned along with the Di Manes on an inscription set up by a certain Artorius, whose name is the original of the Welsh Arthur.

In Istria it may be noted that there is also a goddess Noriceia, who was also called Veica, as we see from the dedication Veicae Noriceiae.

The name Celeia, which is mentioned along with Noreia, is that of the town Cilli in Noricum, a town which was also wor shipped as a goddess, as is shown by inscriptions from the beginning of the third century a.d. On these she is called Celeia Augusta and Celeia sancta. As a goddess she is named on inscriptions along with Noreia in one case and Epona in another.

That Noricum was largely Celtic in its religion may be gathered from the prominence there of the worship of the god Belinos, a name which forms the second element in the British Cuno-belinos, the Welsh Cyn-felyn.

In the course of this investigation it has become very evident how large a part was played in Celtic as well as in at least some forms of Germanic religion by the worship of grouped goddesses. It is from these that the individual goddesses appear in some cases to have been detached, or else developed by a kind of process of unification and generalization. In some cases, topographical connections operated towards individualization, in others the growing conception of the earth as 'the Mother' par excellence, while in other cases the individual goddesses seem to have been the human representatives of previous goddesses of animal form. Of the latter type were doubtless Epona and Damona. In spite of the existence of certain individual goddesses, however, it is most remarkable that the grouped goddesses held their own, especially in certain districts. How far we may base ethno-logical conclusions upon this is very uncertain, since under similar conditions of civilization similar religious ideas are apt to prevail. It is noticeable, however, that the worship of the Matres and the Proximae held their ground even in districts which came under the full influence of Roman civilization.

The Celtic Review III, 1906-1907, Professor E. Anwyl