LIBRARY UNIVERSITY OF CALIFORNIA RIVERSIDE Camtjri^ffe 9[rrf)aeoIogiral anlr a^tftnologiral ^ttits THE HEROIC AGE Longitude 5 E.of Greenw. 10 MAP OF CENTRAL EUROPE illustrating the Heroic Age of the Teutonic Peoples Where the same name occurs both in capitals and italics the former denote a position occupied in the early part of the Heroic Age, while the latter mark a change or extension of territories. THE HEROIC AGE Q-y BY H. MUNRO CHADWICK Fellow of Clare College, Cambridge Cambridge : at the University Press 1912 Reprinted 1967 Published by the Syndics of the Cambridge University Press Bentley House, 200 Euston Road, London, N.W. 1 American Branch: 32 East 57th Street, New York, N.Y. 10022 PUBLISHER S NOTE Cambridge University Press Library Editions are reissues of out-of-print standard works from the Cambridge catalogue. The texts are unrevised and, apart from minor correctionsi, reproduce the latest published edition. TD A/ ^ 'O ^\ -^ First published 1912 Reprinted 1967 First printed in Great Britain at the University Printing House, Cambridge Reprinted in the United States of America Library of Congress Catalogue Card Number: 13-13081 PREFACE THE type of poetry commonly known as heroic is one which makes its appearance in various nations and in various periods of history. No one can fail to observe that certain similar features are to be found in poems of this type which are widely separated from one another both in date and place of origin. In view of this fact it has seemed worth while to attempt a comparative study of two groups of such poems with the object of determining the nature of the resem- blances between them and the causes to which they are due. Occasional illustrations have been taken from other groups of poems belonging to the same type. The first part of the book deals with the early heroic poetry and traditions of the Teutonic peoples, more especially with those stories which were the common property of various Teutonic peoples. It is pointed out that these stories all relate to a period with definite limits — a period for which a consider- able amount of information is available from external sources. The subjects discussed include the distribution of the stories and the relationship between the various versions of them, the antiquity of the earliest poems and the conditions under which they were produced. Lastly, an attempt has been made to estimate the significance of the various elements, historical, mythical and fictitious, of which the stories are composed. The second part deals with Greek heroic poetry and^ traditions. These relate to a period for which little or no external evidence is available; and consequently they present many problems, the bearings of which can hardly be estimated without reference to the existence of similar phenomena else- where. In general I have followed the same plan as in the Vlll PREFACE first part, and made use throughout of the results obtained there. If some excuse is necessary for dealing with so well worn a theme I may plead that, so far as my knowledge goes, it has not hitherto been approached from this point of view. In the third part attention has been called to the existence of a number of somewhat striking characteristics common to the two groups of poems and an attempt made to account for them. The conclusion to which I have been brought is that the resemblances in the poems are due primarily to resem- blances in the ages to which they relate and to which they ultimately owe their origin. The comparative study of heroic poetry therefore involves the comparative study of ' Heroic Ages'; and the problems which it presents are essentially problems of anthropology. In this conclusion I am glad to find myself in agreement with Dr Haddon, who suggested to me that a comparative study of this kind would be a suitable subject for the Cam- bridge Archaeological and Ethnological Series. I take this opportunity of thanking him for bringing the matter before the Syndics of the University Press and for the interest which he has kindly taken in the progress of the work. Owing to the pressure of teaching and other duties a con- siderable time has unfortunately elapsed since the earlier portions of the book were printed. I would therefore respectfully call the reader's attention to the list of Addenda at the end, where references will be found to several important works which have appeared in the meantime. In a work, such as this, which deals with records preserved in a number of different languages, difficulties necessarily arise with regard to the spelling of proper names. In the represen- tation of Teutonic names the system adopted in my previous books has in general been retained. Any such system is of course open to objections, of the cogency of which I am quite aware; and consequently I have not felt inclined to carry out my scheme with rigid consistency. The same remarks apply to the representation of Greek names — which will doubtless displease many critics. South Slavonic names and words are given according to the usual Croatian orthography. PREFACE IX I cannot attempt here to enumerate the various scholars to whose writings I am indebted. It will be seen that they are many and that much of what I have had to do is in the nature of criticism. One name however, that of Professor Ridgeway, I cannot leave unmentioned, since it is largely to his inspiring influence — by no means through his writings alone — that my interest in these subjects is due. It remains for me to record my obligations to a number of friends who have generously responded to my requests for information or criticism on various points. In particular I must mention Miss A. C. Paues, Mr A. B. Cook, Dr W. H. R. Rivers, Mr S. A. Cook, Professor J. W. H. Atkins, Professor A. Mawer, Mr E. H. Minns and Mr F. W. Green. Above all I am indebted to Mr E. C. Quiggin and Mr F. G. M. Beck, who have most kindly read through a considerable part of the book in proof and several chapters even in manuscript. It is scarcely necessary to add that in the sections dealing with Celtic history and poetry Mr Quiggin's criticism has been of the greatest value to me. My thanks are due, further, to my pupils, Mr C. A. Scutt, of Clare College, and Mr Bruce Dickins, of Magdalene College, for similar kind services in the proofs of the later chapters. From the staff of the University Library — in particular I must mention Mr A. Rogers and Mr O. Johnson — I have received the same unfailing and courteous attention as in the past. Lastly, I have to thank the Syndics of the University Press for undertaking the publication of the book and the staff for the efficient and obliging way in which the printing and corrections have been carried out. H. M. C. December, 191 1. CONTENTS CHAPTER I. The early narrative poetry of the Teutonic peoples II. The Heroic Age of the Teutonic peoples III. Scene and nationality in the heroic stories . IV. The origin and history of the heroic poems . V. The poetry and minstrelsy of early times VI. Supernatural elements in the heroic stories VII. Mythical elements in the heroic poems VIII. The use of fiction in the heroic poems IX. The Heroic Age of Greece . X. The Homeric poems XI. Early Greek poetry and minstrelsy . XII. -'Supernatural elements in the Homeric poems XIII. Myth in the Homeric poems J XIV. Fiction in the Homeric poems . . XV. The common characteristics of Teutonic and Greek heroic poetry XVI. Society in the Heroic Age XVII. Government in the Heroic Age .... XVIII. Religion in the Heroic Age XIX. The causes and antecedent conditions of thi Heroic AfiE PAGE I 19 30 41 n no 131 151 168 193 221 249 263 292 320 344 366 393 432 CONTENTS xi NOTES. CHAPTER PAGE I. The use of heroic names in England ... 64 II. On the dating of certain sound-changes in Anglo- Saxon 66 III. Literary influence in Beowulf 73 IV. On the Heroic poetry of the Slavonic peoples . loi V. The Heroic poetry of the Celtic peoples . . 105 VI. The Trojan Catalogue 244 VII. The Battle of Kossovo in Servian poetry . . 313 VIII. The social, political and religious characteristics OF the Celtic and Slavonic Heroic Ages . . 427 Addenda et Corrigenda 464 Index 469 MAPS. Central Europe, illustrating the Heroic Age OF THE Teutonic Peoples .... To face Title-page Greece, illustrating the Catalogue of Ships Greece, showing the distribution of the dia- lects IN historical times between pp. 288 and 289 I TO JAMES RENDEL HARRIS IN GRATITUDE AND FRIENDSHIP CHAPTER I. THE EARLY NARRATIVE POETRY OF THE TEUTONIC PEOPLES. The remains of English poetry which have come down to us from times anterior to the Norman conquest are mainly of a religious character and deal with the lives of saints or with subjects derived from the Bible or ecclesiastical tradition. The secular poems are comparatively few in number and, with one exception, of inconsiderable length. Most of them are narrative poems and admit of a very obvious classification according to- the choice of subjects with which they are concerned. One group deals with the exploits of English kings and noblemen of the tenth century, the other with the exploits and adventures of persons who did not belong to this country. We will take the second group of poems first, as it is admitted by all authorities to be the earlier of the two. The longest poem of this class is Beowulf, an epic of 3183 verses, the subject of which briefly is as follows: Hrothgar, king of the Danes, has built a splendid hall, but is unable to enjoy the use of it on account of the ravages of a monster named Grendel, who attacks the hall by night and devours all whom he finds there. Beowulf, a nephew of Hygelac, king of the Geatas, hearing of Hrothgar's distress comes to his help and destroys first the monster himself and then his mother who had come to exact vengeance. He is thanked and rewarded for his exploits by Hrothgar, and returns to his own home. After this a long period is supposed to elapse. Hygelac has perished in an expedition against the Frisians, and his son Heardred has been slain by the Swedes. Beowulf has succeeded to the throne and 2 EARLY TEUTONIC POETRY [CHAP. reigned many years. In his old age he resolves to attack a dragon which is ravaging the land, and in spite of the cowardice of his followers, of whom all except one forsake him, he eventually succeeds in destroying it, though not before he has himself received a mortal wound. The poem ends with an account of his funeral. The action is interrupted a good many times by references to incidents in the history of the royal families of the Danes and the Geatas, particularly to Hygelac's fatal expedition and to the dealings of his family with the Swedish kings Ongentheow, Onela and Eadgils. We find also a number of allusions to heroes of the past such as Sigemund, Eormenric, Finn and Offa, who are known to us from other sources. To the same class of poetry belong some fragments dealing with the stories of Finn and Waldhere. The fragment relating to Finn is very obscure and indeed would be quite unintelligible were it not for a passage in Beowulf (vv. 1068 — 1159) where the same story is introduced as the subject of a recitation by Hrothgar's minstrel. A certain Hnaef, represented as a vassal of Healfdene, Hrothgar's father, was slain in a fortress belonging to Finn, king of the Frisians. His followers made so brave a defence that Finn was compelled to come to terms with them. Subsequently, when an opportunity presented itself they took vengeance upon Finn and slew him. The fragment gives an account of the fighting which took place, probably in the encounter immediately after Hnaefs death. The story of Waldhere is well known from German sources which we shall have to mention presently. One of the fragments is taken up by an altercation between Waldhere and the Burgundian king Guthhere, before they begin to fight, while the other contains an exhortation to Waldhere by the lady to acquit himself bravely. Widsith, though not an epic poem itself, refers to a large number of the characters which figure in Beowulf, Finn and Waldhere. It is stated that the poet was in the service of a certain Eadgils, prince of the Myrgingas, and that in company with Ealhhild, apparently a princess of the same family, he visited the court of the Gothic king Eormenric. The greater IJ EARLY TEUTONIC POETRY 3 part of the poem is taken up with lists of peoples which he had visited and of famous princes whom he knew personally or by report. The elegy of Deor consists of a number of brief notices of misfortunes which had befallen various persons, Weland and Beaduhild, Geat, Theodric and the subjects of Eormenric. Each notice ends with a refrain expressing the belief or hope that the poet will be able to survive his misfortunes as they did. At the end he says that he had been the bard of the Heodeningas, but that his office had been taken away from him and given to a skilful minstrel named Heorrenda. In a later chapter we shall have to discuss the question when these poems were composed. At present it will be enough to remark that though the MSS. in which they are preserved date only from the tenth or eleventh centuries almost all scholars agree that the poems themselves cannot be later than the eighth century, while probably the majority would refer their composi- tion, in part at least to the seventh. In their present form they cannot be earlier than this,, for with the exception of the Finn- fragment all of them contain Christian allusions. The later group of secular narrative poems may best be described as historical. The earliest of them celebrates the battle of Brunanburh won by Aethelstan in 937 over the Scottish king Constantine II and his Scandinavian allies. Others describe Edmund's conquest of the Five Boroughs, the coronation of Edgar, the glories of his reign, the troubles which followed, and lastly the death of Edward the Confessor. The longest of all is a detailed account of the disastrous battle of Maldon, in which Byrhtnoth, earl of Essex, lost his life. All these pieces except the last are found inserted in texts of the Saxon Chronicle and all without exception appear to have been composed soon after the events which they com- memorate. In addition to the above there are a number of other short poems which are not essentially of a religious character. The most important of these are the Wanderer, the Wife's Complaint, the Husband's Message and the Ruin — to which we may per- haps add the first half of the Seafarer. They are probably all of 4 EARLY TEUTONIC POETRY [CHAP. fairly early date, but they differ from the poems we have been discussing in the fact that they contain no proper names. Those of them which can be called narrative deal apparently with typical characters or situations. A certain amount of magical and gnomic poetry has also survived, while metrical riddles are numerous, but these need not concern us here. It is scarcely open to doubt that a large amount of Anglo- Saxon narrative poetry has perished. Several historical poems and ballads of the tenth and eleventh centuries, now lost, can be traced in texts of the Saxon Chronicle and in Latin works ^ Attempts have been made also to show that narrative poems were used by the compilers of the early part of the Chronicle and by several Latin histories referring to the same period, but the evidence adduced is very doubtful. Perhaps the most likely case is the story of Hengest and Horsa, especially in the form in which it appears in the Historia Brittonum. On the other hand it is probable that parts of the Vitae Duorum Offarum, a St Albans work dating from the beginning of the thirteenth century, are derived ultimately from poems which described Offa's single combat and marriage — incidents to which we find brief references in Widsith and Beowulf respectively. A similar origin may perhaps be claimed for Walter Map's story (^De Nugis Curialium, II 17) of Gado (Wada). A few corrupt verses of a poem on this subject, obviously of late date, have been preserved in a Latin homily^. The earlier and later groups of narrative poems have in general* the same metrical form and on the whole a very similar terminology. The love of battle-scenes is also common to both. In other respects however they differ greatly. Here we need only notice the entire difference in subject-matter ; the poems of the second group contain no allusion to the subjects of the first. Much greater changes however, both in form and matter, are noticeable when English poetry reappears in the thirteenth and fourteenth centuries. The majority of these poems are of French 1 Cf. Brandl in Paul's Grundriss d. germ. Pkilol.'^, ii pp. 1083 flf., 1087 f. ■■' Academy, 1896, I 137 ; cf. Brandl, op. cit., p. 1085. 3 This remark applies more especially to the poems on Brunanburh and Maldon in the later group. I] EARLY TEUTONIC POETRY 5 origin. But even when the scene is laid in England, the subjects are usually drawn from written sources — which is certainly not the case with the historical poems of the tenth century. To the subjects of the earlier group of Anglo-Saxon poems there is scarcely a reference. The period of German literature which corresponds chrono- logically to the Anglo-Saxon period in England is far inferior to the latter in remains of secular narrative poetry. We possess only one fragment of a poem somewhat similar to the Finn- fragment, preserved in a MS. dating from about 800, and one poem of somewhat later date, celebrating a victory of Ludwig III, king of the West Franks. To these we may add three or four very short metrical charms, similar to the Anglo- Saxon magical pieces mentioned above. The subject of the first of these poems is as follows : Hilde- brand {Hiltibrant) is an old warrior who has left his country with Dietrich {Tkeotrih) and served with the Huns. On his return from exile, thirty years later, he is challenged by a young warrior named Hadubrand. In the altercation with which the piece opens Hildebrand discovers that his opponent is his own son, acquaints him with the fact and tries to dissuade him from the combat with offers of rich presents. But the young man refuses to believe him, and taunts him with cowardice and guile in trying to put him off his guard. Hildebrand is therefore obliged to fight, and the fragment comes to an end in the midst of the encounter. The Ludwigslied is a poem of fifty-nine verses celebrating the praises of Ludwig III, with special reference to his victory over the Northmen at Saucourt in 881, and seems to have been composed before the king's death in the following year. It is not in the old Teutonic alliterative metre which we find in Anglo-Saxon poetry and in the Hildebrandslied, but in the later rhyming verse. The religious element is prominent throughout. There is no doubt that a considerable amount of secular narrative poetry once existed in German. Einhard in his Life of Charlemagne (cap. 29) states that the emperor collected native and very ancient poems in which were related the deeds 6 EARLY TEUTONIC POETRY [CHAP. and battles of kings of former times\ But during the following centuries poetry of this kind seems to have gone entirely out ol favour among the higher classes. We do indeed find occasional references to such poems in later Latin works. In particular the Annals of Quedlinburg have incorporated from them a number of notices relating to Eormenric, Theodric {Thideric de Berne) and other heroes of antiquity. But here it is expressly stated that it was among the peasants {rustici) that these poems were known. Other Latin chronicles cite lost poems relating to persons and events of the tenth century, which may have been somewhat similar to the contemporary Anglo-Saxon poems, though generally they appear to have been of a less serious character. Some compensation for the loss of early German poetry is afforded by the preservation of a few Latin poems, of which the most important is Walthariiis manu fortis, composed probably about 930 by Ekkehard of St Gall. The subject is the same as that of the Anglo-Saxon Waldhere fragments. Waltharius and Hiltgund, the son and daughter respectively of two princes in Gaul, were betrothed in their childhood, but had to be given up as hostages to Attila, king of the Huns. After many years they escaped and carried off with them much treasure. Hearing of this. King Guntharius, who dwelt at Worms, determined to waylay them and set out with twelve warriors, among them the brave Hagano who had formerly shared Waltharius' exile. Waltharius is overtaken in a defile of the Vosges and slays eleven of the warriors in single combat. In his final encounter with Hagano and Guntharius all three are crippled, but Waltharius is able to make his way home with Hiltgund. There are other poems, one of which combines German with Latin in each verse, celebrating the deeds of the contemporary Saxon emperors and their relatives. The second period of German literature, beginning in the twelfth century, is incomparably superior to the first in the J Saxo Poeta, who wrote about 890, speaks of uulgaria carmina which celebrated Pippinos Carolos Hludowicos et Theodricos et Carlomannos Hlothariosque (Mon. Germ., Script. I 268). But it is usually held that these words are due to a mistaken inference from Einhard, whose work he was using. l] EARLY TEUTONIC POETRY 7 amount of secular narrative poetry which it has left behind. Here we need only concern ourselves with those poems which draw their subjects from ancient native traditions. These are mostly anonymous and come from the southern districts, especially Austria and Bavaria. Their metrical form is modern and similar to that of other poetry of the same period. The best known of these poems is the Nibelungenlied, which dates from the beginning of the thirteenth century. Only a very brief resume of its contents can be given here. Siegfried {Stvrit), the son of Siegmund, comes from Xanten to Worms and asks for the hand of Kriemhild, the sister of the Burgundian king Gunther. He joins Gunther in his campaigns and by magical arts enables him to win the amazon Brunhild {Prilnhilt) for his bride. The two wives quarrel, and Brunhild learns from Kriemhild of the part played by Siegfried towards her. On hearing this she begins to long for his death, and eventually he is murdered at a hunting party by Hagen {Hagene), the chief of Gunther's knights. Hagen also deprives Kriemhild of Siegfried's treasure {der Nibelunge hort) and sinks it in the Rhine. Kriemhild is afterwards married to Etzel, king of the Huns and, burning for vengeance, persuades him to invite Gunther on a visit, together with his brothers Gemot and Giselher and also Hagen. Soon after their arrival she brings about a quarrel, and after a huge slaughter on both sides, Gunther and Hagen, who have lost all their men, are captured by Dietrich von Bern, Etzel's vassal. Kriemhild puts them to death, violating the oath which she had given to spare them, and in anger at this treachery Hildebrand, Dietrich's old retainer, slays her. Closely connected in subject with the Nibelungenlied is the Klage, which describes the funeral of those slain in the fighting with Gunther and the lamentation over them. Here too we may notice the Seyfridslied, though in the form in which it has come down to us it belongs to a much later period. In reality it is clearly a combination of two different ballads, both of which deal with Siegfried's early years. The first relates how he was brought up by a smith and slew a dragon. The second gives an account of another similar adventure — this time with a fiery 8 EARLY TEUTONIC POETRY [CHAP. dragon which had carried off Kriemhild from her father's home. Siegfried kills the dragon and at the same time takes possession of the treasure of certain dwarfs, the sons of Nybling. A number of other medieval poems deal with the adventures of Dietrich von Bern and his knights. Alpharts Tod tells the story of the young knight Alphart, Hildebrand's nephew, and his encounter with two of Ermenrich's warriors named Witege and Heime, by the former of whom he was treacherously slain. Dietrichs Flucht relates how Ermenrich {EnnricJt) was instigated by an evil counseller named Sibeche to plot destruction for his nephew Dietrich. Though the plot fails, Hildebrand and several of Dietrich's other knights are captured and imprisoned, and in order to obtain their release Dietrich is forced to go into exile. In the Rabenschlacht we are told how Dietrich with the help of Etzel set out to recover his inheritance. While he is engaged in battle with Ermenrich, his brother Diether and two young sons of Etzel are slain by Witege. Dietrich rides after Witege to exact vengeance, but Witege disappears in the sea. Virginal describes how Dietrich was imprisoned by a giant in a castle called Muter and rescued by Hildebrand, Witege, Heime and others. Further adventures with giants and dwarfs are related in the Eckenlied, Sigenot and Laurin. In the Rosengarten Dietrich is made to fight with Siegfried. The story of Wolfdietrich and Ortnit has come down to us in several different forms, but the outline may be given briefly as follows. Wolfdietrich, the son of king Hugdietrich, is kept out of his inheritance by his brothers or their guardian Saben on the ground of illegitimacy. He betakes himself to a faithful old knight, Berchtung of Meran, who raises an army to help him. But in the battle that ensues Berchtung and his sons are all slain or captured, and Wolfdietrich has to go into exile. He then marries the widow of a king named Ortnit and destroys the dragon which had killed him. Eventually he succeeds in winning his father's kingdom and releasing his faithful followers. Some elements of this story appear to have been incorporated in the romance of King Rother. The story of the poem Kudrun falls naturally into two parts. In the first Hetel {Hetele), king of Denmark, hears of Hilde, a l] EARLY TEUTONIC POETRY 9 princess of Ireland and desires to marry her. As her father, the fierce Hagen, will not consent, Hetel's warriors Wate, Fruote and Horand carry her off. Hagen pursues them to Denmark, but in the fight which follows he is nearly killed by Wate. Finally a reconciliation takes place and Hilde is allowed to marry Hetel. In the second part Hartmuot and Herwig are suitors for Kudrun, the daughter of Hetel and Hilde. The latter is at length accepted, but Kudrun is carried off by Hartmuot and his father Ludwig {Ludewic). A pursuit follows and Hetel is killed by Ludwig. Kudrun is kept at Hartmuot's home for seven years, harshly treated by his mother Gerlind, because she will not consent to become his wife. At last Herwig and her brother Ortwin with the warriors Wate, Fruote and Horand come and rescue her. Ludwig is killed but Hartmuot is spared at Kudrun's intercession. In conclusion mention must be made of some fragments of a poem dealing with Walther and Hildegund. The subject seems to have been identical with that of Ekkehard's Waltharius. Reminiscences of other ancient stories are occasionally to be found in poems of a romantic character. Among these we may note especially the poem Herzog Friedrich von Schwaben, a portion of which seems to be derived from the story of Weland. The North German dialects have no poetry of this type, except a few ballads dating from much later times, of which the most important is one on Eormenric's death. But a great mass of legend, derived chiefly from North German sources, is pre- served in the Norse work ThiSreks Saga af Bern, which dates from about the middle of the thirteenth century. The characters are for the most part identical with those which figure in the High German epics, Dietrich von Bern, Ermenrich, Witege, Walther, Siegfried etc. Traces of poems of much earlier date have been sought in several Latin chronicles. Besides the references in the Annals of Quedlinburg, to which we have already alluded, mention may be made of the Saxon stories given by Widukind and the Translatio S. Alexandri, especially that of the victory over the Thuringian king Irminfrith. The vernacular poetry of the Langobardi has entirely perished, but a number of stories given by the Latin historians lO EARLY TEUTONIC POETRY [CHAP. are thought to be based on early poems. The first and most striking of these is the account of the battle with the Vandals, in which the two armies appeal for victory to Wodan and Fria respectively \ A similar origin has been claimed for several other narratives, such as that of Alboin's visit to Turisind, king of the Gepidae, and more especially the story of Authari and Theudelinda^ The Gothic historian Jordanes states (cap. 5) that his com- patriots were wont to celebrate the deeds of their famous men in poetry, and it is probable that many of the legends which he gives were ultimately derived from such poems. Among them we may include not only the story of Filimer and the migration to the Black Sea (cap. 4) — in which case ancient poems are expressly mentioned — but also perhaps that of the first migra- tion under Berig {ib.) and some part of the account of Ostrogotha (cap. 16 f.), as well as the incidental reference to Vidigoia's death (cap. 34). In the story of Hermanaricus also several incidents, notably the death of Sunilda and the vengeance subsequently exacted by her brothers Sarus and Ammius (cap. 24), suggest a tradition preserved in poetic form. No Scandinavian country except Iceland* has preserved any early poems in its native language. The ancient literature of Iceland however is peculiarly rich in secular narrative poetry. Moreover though the earliest of these poems are probably quite two centuries later than Beowulf, they are entirely free from Christian influence. Indeed it can hardly be doubted that a considerable proportion of them date from heathen times. It will be convenient to begin with the collection of poems usually known as the Older Edda. These are all anonymous ; but most of them are generally believed to belong to the tenth century, while a few may really have been composed in Norway at a still earlier date. Eleven of these poems deal exclusively, or almost exclusively, with gods, giants and other supernatural 1 Origo Gentis Langobardorum (Mon. Germ., Script. Rerum Langobard., p. 2f.); Paulus Diaconus, Hist. Lang, i 8. ^ Cf. Kogel, Geschichte d. deutschen Litteratur, i p. 115 ff. * The poems were not all composed in Iceland. Many of the earliest doubtless came from Norway, others perhaps from the British Isles, while others again are referred to Greenland. I] EARLY TEUTONIC POETRY II beings, and hence stand quite apart from the class of poetry with which we are concerned. But since we shall have to refer to them occasionally in the following pages it will be convenient here to give a brief synopsis of their contents. Voluspd is a mythological poem in the form of a speech delivered to the god Othin by a sibyl whom he is consulting. It deals with the origin of the world, the history of the gods and their coming fate. Hdvamil is a collection of proverbial wisdom and moral precepts of the heathen age, with occasional references to myth and ritual. Vaf}?ruSnismdl describes how Othin visited the giant VafJ^ruSnir and entered into a contest with him in mythological lore. The subject of Grimnismal is a visit paid by Othin in disguise to a king named GeirroSr, who tortures him. Othin gives a long discourse on mythological matters and finally reveals himself, whereupon the king stumbles over his sword and dies. Skirnismdl relates how Skirnir, the servant of the god Frey, was sent to obtain for his master the hand of the giantess GerSr, of whom he had become passionately enamoured. HarbartSslioS is taken up with an altercation between the god Thor and a ferryman called HdrbarSr (generally supposed to be Othin), who refuses to take him over a strait. H^^miskviSa gives an account of Thor's adventures when he went to visit the giant Hymir. Lokasenna is occupied with a number of scandalous charges brought by Loki against various gods and goddesses who have been invited to a feast by Aegir. ThrymskviSa relates how the giant Thrymr, having obtained possession of Thor's hammer, demanded the goddess Freyia as his bride, and how Thor came disguised as Freyia and slew the giant. Alvissmal is a dialogue between Thor and the dwarf Alvi'ss, who is a suitor for his daughter. Thor detains the dwarf with questions on the various names of natural objects and phenomena until the fatal moment of daybreak. VegtamskviSa describes how Othin went to consult a sibyl on the impending fate of Balder. In addition to these pieces, all of which probably come from one collection, there are two semi-mythical poems contained in other works. Rigsmdl or Rigs)?ula relates how a certain Ri'gr (identified in the introduction with the god Heimdallr) became 12 EARLY TEUTONIC POETRY [CHAP. the progenitor of the three classes of men — the characteristics of which are described at length. Hyndluli6t5 is mainly a genea- logical poem, narrating how Freyia went to consult the giantess Hyndla as to the ancestry of her devotee, 6ttarr the son of Innsteinn. It contains also some purely mythological matter which is generally supposed to have come from a separate poem. Here also we may mention two pieces known as Grogaldr and Fiolsvinnsmil, which clearly belong together. In the former Svipdagr calls up the spirit of Gr6a, his mother; in the latter he comes to the enchanted abode of MengloS and, after many questions with the gatekeeper, is at length recognised as her destined lover. In much later times we find Swedish and Danish versions of the same story, and there can be little doubt that it is really a folk-tale. The purely mythological poems of the Edda are followed by VolundarkviSa, which gives a fairly full account of the story of Volundr (Weland). It describes how Volundr and his brothers obtained as their wives three swan-maidens, who after eight years deserted them. Then Volundr is captured by a king named NiSutSr, hamstrung and compelled to work as his smith. Volundr executes vengeance on the king's sons and daughter (BoSvildr) and then flies away. The next poem, HelgakviSa Hiorvart5ssonar, gives in dialogue form a somewhat complicated story of the adventures of a king named HiorvarSr and his son Helgi. A different Helgi, the son of Sigmundr, is the subject of the two following poems, Helga- kviSur Hundingsbana. Both poems relate how the hero over- threw a king named Hundingr and how he was afterwards summoned by Sigrun, the daughter of Hogni, to save her from marriage to a prince named HoSbroddr, whom she detested. The first poem ends with Helgi's victory over Hogni and HoSbroddr ; but the second goes on to describe how Dagr, the son of Hogni, subsequently slew Helgi in revenge for his father. Sinfiotli, the son of Sigmundr, figures in both poems, but SigurtSr is not mentioned. The next poem, Gn'pisspd, probably a late work, gives a summary of the adventures of SigurSr, the son of Sigmundr, in the form of a dialogue between the hero and his uncle Gri'pir, I] EARLY TEUTONIC POETRY 1 3 who is endowed with prophecy. This is followed by three pieces which may really be parts of one original poem. The first (commonly called Reginsmal) relates how three of the gods, Othin, Hoenir and Loki, killed a certain Otr, the son of Hrei?5marr. Having to pay compensation to the father, Loki robbed a dwarf, named Andvari, of his gold. Andvari laid a curse on the gold, and HreitJmarr was soon killed by his own son P'afnir, who subsequently turned into a serpents Reginn, Fafnir's brother, betook himself to SigurSr and became his attendant. The poem then goes on to describe how a certain Hnikarr (Othin) guided SigurSr on an expedition he undertook against the sons of Hundingr. In the next piece (Fafnismal) SigurSr at Reginn's instigation attacks and kills Fafnir. Then, finding that Reginn is plotting treachery, he slays him also and carries off Fafnir's gold. In the third (Sigrdrifumal) SigurSr finds and wakens a maid named Sigrdrifa, a valkyrie who has been punished by Othin with an enchanted sleep. She imparts to him much magical lore ; but the close of the poem is lost owing to a lacuna in the MS. Of the following nine poems six — SigurSarkviSa I (a frag- ment), GuSrunarkviSa I, SigurSarkviSa II {liin skamma), GuSrii- narkviSa II {hiti fornd), AtlakviSa and Atlamal — deal with practically the same events as the Nibelungenlied. But unfor- tunately, owing to the lacuna in the MS, several poems at the beginning of the series have been lost. Hence in order to obtain a full account of the story it is necessary to refer to the Volsunga Saga (see below), which used the lost poems as well as the others. The chief variations from the German version of the story are as follows : (i) Sigmundr has been killed before the birth of SigurSr (Siegfried) by the sons of Hundingr. (ii) The wife of SigurSr is called GuSriin, Grimhildr being the name of her mother and Giiiki that of her father, (iii) Brynhildr is the sister of Atli (Etzel). In several authorities^ she is identified with the valkyrie ^ Called a dragon {dreki) in Volsunga Saga. ^ HelreitS Brynh. (str. 7 ff.) and the prose versions (Skaldsk. 41, Vols. S. 20). It is a much debated question whether the two are identified also in Fafnismal, str. 40 ff. In Gripisspa they seem clearly to be separated. 14 EARLY TEUTONIC POETRY [CHAP. of Sigrdri'fumdl, and it is stated that she and SigurtSr had ex- changed vows of love ; but Sigurt5r's love was subsequently- turned to Gu?5run through a magic potion given him by Gn'm- hildr. (iv) Hogni (Hagen) is the brother of Gunnarr (Gunther); and in place of Gemot and Giselher there is another brother named Guthormr. It is the last-named who actually kills SigurSr, and he is himself killed by the dying man. Brynhildr then puts herself to death and is burned with SigurSr. (v) The death of Gunnarr and Hogni is attributed to Atli. GuSriin on the other hand warns her brothers of the treachery awaiting them, and subsequently avenges their deaths by killing Atli and the children he had had by her. Interspersed among these poems are three others connected with the same story, but dealing with incidents unknown to the German version. The HelreiS Brynhildar describes how Bryn- hildr on her way to Hell encounters a giantess, to whom she tells the story of her life. GuSrunarkviSa HI relates how GuSrun was accused to Atli of adultery with ThioSrekr, but established her innocence by the ordeal. In Oddrunargrdtr a sister of Atli named Oddrun comes to Borgny, the daughter of a certain Heit5rekr, to relieve her in her travail, and gives an account of the relations between herself, Brynhildr and Gunnarr. The last two poems — GuSrunarhvot and HamSismal — are concerned with the story of the attack upon lormunrekr (Her- manaricus), to Jordanes' account of which we have already alluded (p. lo). Here the story is connected with the preceding poems^ by the fact that Svanhildr (Sunilda) is represented as the daughter of GuSrun and SigurSr, while HamSir and Sorli (Ammius and Sarus) are said to be the sons of GuSrun and a certain lonakr, whom she had married after Atli's death. In the first poem GuSrun incites her sons to avenge their sister and then bewails her many misfortunes. The second describes the actual fighting, preceded however by the account of a quarrel in which Erpr, a stepson of GuSrun, is killed by his half- brothers. To his death is attributed the fact that lormunreky'' 1 The death of Svanhildr is mentioned also in Sigur'SarkviSa II, where it is attributed (as in the prose authorities) to the evil counsel of a certain Bikki, an adviser of lormunrekr. I] EARLY TEUTONIC POETRY 1 5 was able to survive the onslaught, though he lost both hands and both feet. The prose Edda (Skaldskaparmal, cap. 43) contains a poem, Grottasongr, which gives the story of two giant maidens who had to grind gold and peace for the Danish king Fr6t5i. At the end there is an allusion to Halfdan and Hrolfr Kraki (see below). Besides this we have, both in the prose Edda and elsewhere, a number of fragments of poems, some of which refer to SigurSr and his family and some to other stories. Among the latter mention may be made especially of a poem Biarkamdl, which celebrated the achievements of the Danish king Hrolfr Kraki and the heroism of his retinue in the battle wherein he lost his life. In Hervarar Saga large portions of an early poem relating to the Goths and Huns have been preserved. In addition to the poems enumerated above, all of which are anonymous, we have also a number of works by known poets. Most of these deal either wholly or in part with contemporary persons and events. The earliest date from the ninth century and are of Norwegian origin. Probably the oldest of all are the fragments of Bragi Boddason, the chief of which is a description of his shield (Ragnarsdrdpa). From the reign of Harold the Fair-haired several poems are known, though nearly all of them are in a very fragmentary condition. Among them may be men- tioned especially the Ynglingatal of ThioSolfr of Hvi'n, a genea- logical poem which traces the descent of Rognvaldr, a cousin of Harold, from the ancient kings of the Swedes and the god Frey. Another famous work by the same poet was the Haustlong, which dealt with mythological subjects. Next perhaps in importance to ThioSolfr was Thorbiorn Hornklofi, from whom we have fragments of two poems (Hrafnsmdl and Glymdrdpa), celebrating the exploits of Harold. Somewhat later we hear of a poet named GoSSormr Sindri, who is known chiefly from the remains of a work (Hdkonardrdpa) in honour of Haakon I. A famous fragment (Eireksmil) by an unknown poet, dating from shortly after the middle of the tenth century, celebrates the death of King Eiri'kr BloSox and his reception by Othin in Valhalla. From this is copied the Hakonarmdl of Eyvindr Skaldaspillir, celebrating the death of Haakon I at the battle of l6 EARLY TEUTONIC POETRY [CHAP. Fitje (a.d. 961). The same poet also composed a genealogical poem (Haleygiatal) in imitation of Ynglingatal. In it he traced the ancestry of Haakon, earl of Lade, who ruled Norway from about 975 to 995, back through the kings of Hdlogaland to Othin and SkaSi. From this time onwards the cultivation of poetry seems to have been almost entirely limited to Icelanders, many of whom resided largely at the courts of various Scandinavian kings. Among them the most noteworthy and almost the earliest was Egill Skallagn'msson, who lived from about 900 to 982. He is known chiefly from the HofuSlausn, composed for Eirikr BloSox in England, the Arinbiarnardripa, in honour of his friend Arinbiorn, and the Sonatorrek, an elegy over one of his sons. Of his younger contemporaries perhaps the best known are Kormakr Ogmundarson and Einarr Helgason. The latter is famous chiefly for his poem Vellekla, in which he celebrated the exploits of Earl Haakon. Many other distinguished poets flourished during the following half century — down to the time of Harold III (HarSrdtSi) — but it is not necessary here to discuss their works. By the time of Harold III the composition of prose narra- tives or sagas (sogur) had already begun to be cultivated by Icelanders, though it was not until towards the end of the following century that they were first committed to writing. Many sagas are based on old narrative poems ; as for instance Ynglinga Saga, which is largely a paraphrase and expansion of Ynglingatal. Volsunga Saga, which gives the stories of Helgi, SigurtJr, GutSriin and Svanhildr, is derived mainly from the poems of the Edda, though it has used other materials. The earlier part, dealing with Sigmundr and his ancestors seems to have drawn upon some lost poems. In Sbrla Thattr^ we meet with a story — found also in the prose Edda (Skaldsk. 50) and alluded to in Bragi's Ragnarsdrapa — which is clearly connected with that of the first part of the German poem Kudrun. Hildr, the daughter of Hogni, is carried ofi" in her father's absence by his friend HetSinn. They are overtaken and a battle follows, in which all the combatants are killed. Hildr by magic spells ^ Fomaldar Sogur Not'Srlanda, i p. 391 ff. l] EARLY TEUTONIC POETRY 1 7 rouses the slain each night to renew the battle. In Hr61fs Saga Kraka we have an account of the Danish kings Helgi and Hroarr, the sons of Halfdan, and of Helgi's son, Hr61fr Kraki, which seems to be derived ultimately from old poems like Biarkamal. Certain incidents in the story, such as Hrolfr's dealings with the Swedish king ASils, are related also in other sagas. The same characters figured prominently in Skioldunga Saga, of which we have little except an abridged Latin trans- lation. This saga also related at length the stories of Haraldr Hilditonn, Sigurt5r Hringr and Ragnarr Lot5br6k, parts of which are known also from other sources, especially Ragnars Saga LoSbrokaj and the Thdttr af LoSbrokar sonum. Among Latin authorities the most important is the great 7 Danish History {Gesta Danornni) of Saxo Grammaticus, which dates from the end of the twelfth century and contains metrical translations or paraphrases of many old poems. Of these perhaps the most noteworthy are the Biarkamdl and some of the poems attributed to Starkat5r (Starcatherus), particularly those ad- dressed to Ingialdr (Ingellus). Here also we find a detailed account of the tragic story of HagbarSr and Signy, which is very frequently alluded to in Old Norse poetry. HagbartSr belonged to a family which was involved in vendetta with the Danish king Sigarr. But having fallen in love with Signy, the king's daughter, he visited her disguised as a woman — an adven- ture which ended in his being discovered and condemned to death. Signy and all her maidens destroyed themselves when he was led to the gallows. Many other stories are given entirely in prose, among them those of HeSinn and Hogni, lormunrekr and Helgi Hundingsbani, though no mention is made of Sig- mundr or Sigurt5r. The adventures also of Haraldr Hilditonn and Ragnarr LoSbrok are related at considerable length. Lastly, we may mention a story which is not recorded by any Icelandic authority, namely that of the single combat fought by Uffo the son of Wermundus. It deserves notice here on account of its obvious identity with the English story of Offa. In the course of this chapter we have reviewed briefly the secular narrative poetry produced by the various Teutonic peoples down to the end of the tenth century. In the case of 1 8 EARLY TEUTONIC POETRY [CHAP. I works by historical Norwegian and Icelandic poets, owing to the abundance of material, we have restricted ourselves to men- tioning only the leading names. Elsewhere we have endeavoured to give a more or less complete summary. On the other hand, among works dating from later than the tenth century, we have taken into consideration only those which are concerned with stories of ancient times. The stories themselves will be discussed in the following chapters. CHAPTER II. THE HEROIC AGE OF THE TEUTONIC PEOPLES. It will not have escaped notice that a large proportion of the stories described or alluded to in the preceding chapter are found in the literature of more than one nation. The most casual reader could not fail to observe the identity of the story of SigurSr and GuSrun, as given in the Edda and Volsunga Saga, with that of Siegfried and Kriemhild related in the Nibelungenlied. Equally obvious is the connection between the story of lormunrekr and Svanhildr in the same Scandinavian authorities and that of Hermanaricus and Sunilda given by Jordanes. A still closer resemblance is furnished by the Anglo- Saxon poem Waldhere and the German-Latin Waltharius. The connection between the German poem Kudrun and the Scandinavian story of HetSinn and Hogni is perhaps less striking, but not open to question. The brief references to the story of Weland and Beaduhild in the Anglo-Saxon poem Deor are quite sufficient to prove its substantial identity with that told in VolundarkviSa. In other cases the same characters appear, though the incidents related are different. Eormenric (Ermenrich) is a prominent figure in Anglo-Saxon and German poetry, as well as in Scandinavian and Gothic records. The Wudga and Hama of Widsith (Waldhere and Beowulf) and the Theodric of Waldhere are clearly identical with the Witege, Heime and Dietrich von Bern of the German epics, while Theodric figures also, though not prominently, in the Edda. Of all the Edda poems^ those which show the least connection with non-Scandinavian poetry are the three HelgakviSur. Yet Sinfiotli, as well as Sigmundr, ^ Except of course those which deal only with supernatural beings (p. ii f.). 20 THE HEROIC AGE [CHAP. is mentioned in Beowulf (under the form Fitela) and there are traces that his name was once known in Germany. We may observe that in Beowulf it is only the persons mentioned in casual references and in episodes lying outside the main action of the story, such as Sigemund, Eormenric, Hama and Weland, to whom we find allusions in German poetry. On the other hand most of the chief characters of the poem are well known from Scandinavian records, though not in connection with precisely the same incidents. There is no doubt as to the identity of the Danish kings Healfdene, his sons Hrothgar and Halga, and Hrothwulf the nephew and colleague of Hrothgar, with Halfdan, his sons Hroarr and Helgi and Helgi's son, the famous Hrolfr Kraki — all likewise kings of the Danes. Equally obvious is the identity of the Swedish prince Eadgils the son of Ohthere with ASils the son of Ottarr in Ynglingatal, while his uncle and opponent Onela is clearly the same person as A?5ils' opponent AH, although the latter is represented as a Norwegian in Old Norse literature. Further, the episode in which Ingeld is incited by an old warrior to avenge his father Froda is evidently to be connected with certain poems given by Saxo, in which the old warrior Starcatherus rouses Ingellus to avenge his father Frotho. Among other persons mentioned in the poem Scyld is doubtless to be identified with the Skioldr of Scandinavian tradition ; probably also Heremod with the HermoSr of Hynd- lulioS and Weoxtan with the Vdsteinn of the Kalfsvi'sa\ To the identification of Beowulf himself with the Biarki of Scandinavian tradition, which is doubted by some scholars, we shall have to return later. The characters of the Finn fragment are much less easy to trace elsewhere. Two of Hnaef's warriors named Ordlaf and Guthlaf are probably to be identified with two Danish princes, Oddlevus and Gunnlevus, mentioned in Skioldunga Saga ; but there is nothing to show that Sigeferth is identical with SigurSr the son of Sigmundr. In Widsith however we find a large number of persons who are well known from Continental and Scandinavian authorities. Besides Eormenric, Wudga, Hama, Offa, Hrothgar, Hrothwulf and Ingeld, of whom we have already spoken, ^ Quoted in the prose Edda (Skaldskaparmal, cap. 66). II] THE HEROIC AGE 21 we hear of Aetla, king of the Huns (Atli, Etzel), Guthhere and Gifeca, kings of the Burgundians (i.e. the Gunnarr and Giuki of the Edda), and Sigehere, king of the Danes (i.e. probably Sigarr, the father of Signy), as well as the Goths Eastgota the father of Unwine (i.e. Jordanes' Ostrogotha the father of Hunwil), Becca (probably Bikki, the evil counsellor of lormunrekr) and the Herelingas, Emerca and Fridla, who are doubtless to be identified with the Embrica and Fritla, nephews of Eormenric, mentioned in the Annals of Quedlinburg and elsewhere. It is more than probable also that in v. 21 : " Hagena ruled the Holmryge and Heoden (MS. Henden) the Glommas," we have an allusion to the story of HetJinn and Hogni (the Hetel and Hagen of Kudrun). We may compare a passage of Deor (v. 35 ff.) which tells of a skilful minstrel of the Heodeningas -named Heorrenda, pre- sumably the Horand of the German poem. These instances, though far from exhaustive, will be sufficient to show that the same characters recur again and again in the early narrative poetry of the various Teutonic peoples. In the last nineteen poems of the Older Edda (viz. those which deal with human beings) there is but one (Helgakvit5a Hiorvart5ssonar) which introduces no characters known elsewhere. Among the early Anglo-Saxon poems treated on p. i ff. we find no such case, and the same is probably true of the German poems discussed on pp. 5, 7 ff. With the later Anglo-Saxon poems (p. 3) and the German historical poems of the ninth and following centuries the case is quite otherwise. The exploits of Aethelstan and Byrhtnoth are celebrated only in English poems, those of Ludwig III and other German princes only in the poetry of their own country. So also with the skaldic poems of the North. If these introduce any personal names known in the poetry of England or Germany they are names, like Sigmundr and HermoSr, derived from earlier poems and not belonging to contemporary persons. Similarly neither English nor German poetry celebrates the deeds of Eiri'kr BloSox or Haakon the Good, The phenomena noted above seem to indicate that the poetic cycles with which we have been dealing have a common origin or at least that there was a considerable amount of borrowing 22 THE HEROIC AGE [CHAP. between poets of different nations. In order however to be able to form an opinion on this point it is necessary first to consider the following questions : (i) how far the characters and incidents of these poems are to be regarded as historical, (ii) to what period or periods of history they belong. We may note in passing that no doubt need be entertained as to the historical basis of the later group of Anglo-Saxon poems, of German poems such as the Ludwigslied or of the Northern skaldic poems — however much the true facts may be obscured by poetic embellishments. It is clear enough that some of the characters of the common cycles are historical persons. Thus there can be no doubt that Aetla (Atli, Etzel), king of the Huns, is the famous Hunnish king Attila who died in 453. Again the Burgundian king Guthhere (Gunnarr, Gunther), who plays so prominent a part in the stories of Waldhere and SigurSr-Siegfried, is clearly identical with the historical Burgundian king Gundicarius (Gundaharius), whose defeat in 435 by the Roman general Aetius is recorded by con- temporary writers. Of his end Prosper says only that the Huns destroyed him together with his family and nation \ and some scholars have denied that Attila had any part in this eventl But our knowledge of the course of events on or beyond the Roman frontier at this period is too slight to justify any confident state- ment on such a point. Of the other members of the Burgundian royal family Gifeca (Giiiki) and Gislhere (Giselher) are mentioned in the laws of King Gundobad who died in the year 516^ ' Theodosio xv et Valentiniano iv coss. (a.d. 435): pax facta cum Vandalis... eodem tempore Gundicarium Burgundionum regevi intra Gallias habitantem Aetius bello oblriuit, pacemque ei supplicanti dedit, qua non diu potitus est. siquidem ilium Hunni cum populo atque stirpe sua deleuerunt. ^ The overthrow of Guthhere is ascribed to Attila by Paulus Diaconus in his Gesta episc. Mettensium (Mon. Germ., Scr. \\ p. 262), Hist. Misc. XIV (Muratori, Scr. I p. 97) ; but he is a late authority. On the other hand an anonymous Gaulish chronicle, which ends in the year 452, attributes the whole of the Burgiuidian disasters to Aetius : bellum contra Burgundionum gentem memorabile exarsit quo uniuersa pene gens cum rege per Aeiium deleta (Mon. Germ., Chron. Min. I 660). But this seems to be due to the confusion of two events which Prosper clearly distinguishes (cf. Idatius, Chron., Theodosii ann. Xii, XIll). * Liber Legum Gundebati, cap. 3 (Mon. Germ., Leg., Vol. III p. 533): si quos apud regiae memoriae auctores nostras, id est Gebicam, Godomarem, Gislaharium, Gundaharium, patretn quoque nostrum et patruum, liberos liberasue fuisse consliterit, in eadem libertate permaneant. it] the heroic age 2$ The Gothic king Eormenric (Hermanaricus, lormunrekr) is another doubtless historical character. The account of him given by Jordanes seems indeed to be derived from tradition, handed down probably in poetic form ; but the statement that he took his own life through fear of the Huns is confirmed by the strictly contemporary writer Ammianus Marcellinus (xxxi. 3. i), from whom we gather that the event took place shortly after 370. He also states that Eormenric was a most warlike king and feared by the surrounding nations on account of his many brave deeds. In Dietrich von Bern (the Theodric of Waldhere and the ThioSrekr of the Edda) we certainly have reminiscences of the Ostrogothic king Theodoric who ruled Italy from 489 to 526. The statement in the Hildebrandslied that he fled from the hostility of Ottachar and the story of the Rabenschlacht recall his campaigns with Odoacer, which culminated in the surrender of Ravenna in 493. But it cannot possibly be true that he was present at Attila's court, where we find him in the Edda and in German poetry, much less that he had any dealings with Eormenric. The former mistake is generally attributed to confusion between Dietrich and his father Dietmar (Thiudemer), who is known to have been with Attila. The other error however is more difficult to account for and will require to be discussed later. It is commonly held that in the poems dealing with Wolfdietrich^ the hero and his father Hugdietrich represent a confusion of the Prankish kings Theodberht (r. 534 — 548) and his father Theodric I (r. 511 — 534). In this case however it must be confessed that the resemblances are extremely slight. The application of the name Hugo Theodoricus to Theodric I in the Annals of Quedlinburg cannot at best prove more than that the chronicler identified the two. The identification of characters which figure in stories relating to the northern kingdoms is naturally more difficult, since references to such persons by contemporary Roman 1 This cycle is supposed to have been known in England at one time, owing to the juxtaposition of the names Seafola (Saben) and Theodric in Widsith, v. 115. Theodric, king of the Franks, is mentioned in the same poem (v. ■24). t 24 THE HEROIC AGE [CHAP. historians are extremely rare. One safe instance however is furnished by the incident, referred to several times in Beowulf, of Hygelac's disastrous expedition against the Franks and Frisians. Gregory of Tours (ni 3) and the Gesta Francorum (cap. 19) mention a very serious raid on the lower Rhine by a king of the Danes named Chocilaicus, which ended in his defeat and death through the arrival of an army under Theodberht. The Liber Monstrorum (I, cap. 3^), a work of perhaps the seventh century, states that the bones of a certain Getarum rex Huiglaucus, who had been slain by the Franks, were preserved on an island at the mouth of the Rhine. There can be no question that the person referred to in these passages is the Hygelac of Beowulf The date of the expedition, though not precisely fixed by any authority, may safely be placed within a few years of 520. Most of the Danish and Swedish princes common to Beowulf and the Northern authorities are now generally regarded as historical characters, though we have no reference to them in contemporary documents. It is to be noted in the first place that though the persons themselves are common to the two traditions, English and Northern, they are not as a rule mentioned in connection with the same incidents. Further, there is no evidence for communication between England and the Baltic during the seventh and eighth centuries. This renders it probable that the two records go back independently to a time at which persons who remembered Hygelac's younger contemporaries might still be alive. Lastly a few words must be said with regard to the stories of Haraldr Hilditonn, SigurtJr Hringr and Ragnarr Lot5br6k. The sons of LotJbrok are well known from contemporary historical documents through their piratical expeditions, more especially the great invasion of England in 866. Moreover, though the references to Lot5br6k himself are rare and doubtful, it is clear enough that the king Ella who is said to have put him to death was the Northumbrian usurper Aella, who reigned from 863 to 867. Loc5br6k's father SigurSr Hringr has been identified with a certain Sigifridus whose conflict with another Danish king named Anulo is recorded in a number of Latin 1 Cf. Berger de Xivrey, Traditions T^ratologiques, p. 1 2. II] THE HEROIC AGE 2$ chronicles under the year 812, and Anulo himself with that Oli who is represented as Sigurt5r's ally at the battle of Brdvi'k. Lastly, it has been suggested that a reference to Haraldr Hilditonn may quite possibly be preserved in the description of Anulo as nepos Herioldi quondam regis^ ; for according to Saxo (p. 250) Oil was the son of Haraldr's sister and eventually succeeded him on the Danish throne. The above identifications^ are sufficient to show that historical I characters are introduced into most of the stories with which we have been dealing. Further — and this is a very remarkable fact — apart from the last cycles embracing Haraldr Hilditonn and Ragnarr LoSbrok, which are entirely confined to Northern literature, all the historical personages whom we have been able to ident ify belong to a period extending over barely two centuries. Eormenric flourished in the latter half of the fourth century, Attila and the Burgundian kings in the first half of the fifth ; Theodric towards the end of the same century and in the first quarter of the sixth. Hygelac again was a contemporary of Theodric ; while Wolfdietrich, if he is rightly identified with Theodberht, died in 548. In the stories which form the common themes of English, German and Scandinavian poets we find no mention of historical persons who lived after the middle of the sixth century. Now it will be clear that the cycles of stories dealing with Ragnarr LoSbrok and his ancestors are really, like the skaldic poems, to be compared with German and English works such as the Ludwigslied and the poem on the battle of Brunanburh. The difference in tone is sufficiently accounted for by the social conditions of the Viking Age, which were wholly different from those which prevailed in the Christian kingdoms. The statement that the common cycles of tradition mention no historical characters later than about 550 ought perhaps to be qualified in one case. Paulus Diaconus {Hist. Lang, l 27) says ^ Einhardi Ann., 812 (Mon. Germ., Scr. I p. 199). Prof. Olrik {Nordisches Geistesleben, p. 44) apparently rejects this identification, as he places Haraldr Hilditonn not long after the time of Hrolfr Kraki. ^ The list makes no claim to completeness. Thus several of the characters in Hervarar Saga (probably mentioned also in Widsith, v. ii6fif.) have been identified with historical persons of the fifth century. But the evidence is far from satisfactory. 26 THE HEROIC AGE [CHAP. that the praises of Alboin, king of the Langobardi, who died in 572 (or 573), were sung by the Saxons, Bavarians and other peoples; and it has been suggested that his account of certain incidents in Alboin's career is derived from poetic sources. Further, we find the generosity of the same king celebrated in the Anglo-Saxon poem Widsith, where he is apparently the latest person mentioned. Hence there is some ground for including him among the characters of common Teutonic poetry — which will involve our extending the lower of the chronological limits fixed above by about twenty years. Yet it is not clear that Alboin figured in any poems which can properly be called narrative, except perhaps among his own people. On the other hand some of the Gothic heroes recorded by Jordanes, if we are to trust his chronology at all, must be referred to times long anterior to the middle of the fourth century. Apart from Gothic tradition the only mention of any of these persons occurs in a brief passage in Widsith (v. 113 f.) : "(I have visited) Eastgota, the wise and good father of Unwine." Now Ostrogotha is brought by Jordanes into connection with the Emperor Philip (v. 244-^249). He is mentioned also by Cassio- dorus ( Var. XI. i) as one of the ancestors of Amalasuintha and as a prince renowned for forbearance (patientia). In spite of his suspicious name what is said of him by Jordanes seems to point to a genuine tradition. But if so, even setting aside both the reference to Philip and the genealogy given by Jordanes in cap. 5, he is probably to be referred to a time anterior to the upper limit fixed above. Into the story of Filimer and the migration we need not enter, as there is no reason for supposing it to be anything but a purely Gothic tradition. The story of the Langobardic victory over the Vandals (cf p. 10) is probably of a similar character. There remain of course a number of stories which contain no names of persons mentioned in contemporary records. The story of Finn is in Beowulf connected, rather loosely, with Healfdene, Hrothgar's father. Hence if the incidents which it relates are to be regarded as historical, they must be dated somewhat earlier, though certainly not more than a century earlier, than Hygelac's expedition. The story of Ofifa and his II] THE HEROIC AGE 27 father Wermund must be referred to a still earlier period if we are to trust the evidence of the M-ercian genealogy in which these persons figure. I have tried elsewhere' to show that the Athislus, who in Saxo's version of the story appears as Wer- mund's enemy, is probably to be identified with the Eadgils prince of the Myrgingas mentioned in Widsith. The latter is represented as the contemporary of Eormenric, and the date thus obtained agrees with that given by the genealogy. The story of Weland, if it contains any historical element, should be placed perhaps slightly further back ; for in Waldhere, as well as in many German authorities, including ThiSreks Saga, Weland is said to be the father of Widia ( Wudga, Witege). The latter is often associated with Eormenric-, and there can be little doubt that he is to be identified with the Gothic Vidigoia who is mentioned as a hero of the past by Jordanes (cap. 34) in a quotation from Priscus^'. For the story of Hagbart5r and Signy a date is afforded by a poem attributed by Saxo (p. 214) to StarkaSr. The poet, who in his old age served Ingellus, i.e. Ingeld, the son-in-law of Hrothgar, says that he had followed Haki, the brother of HagbarSr, in his early youth. In Ynglinga Saga (cap. 23) Haki is made to fight with the Swedish king lorundr, four generations above ASils, but this genealogy cannot be entirely correct. The only important stories which remain are those of FroSi the Peaceful and HeSinn and Hogni. In Saxo's history (p. I58ff.) the two are brought into connection with one another, and it is certainly to be noted that a Fruote von Tenemarke plays rather a prominent part in Kudrun, especially the first portion. But FroSi is associated with different sets of persons in different works, and his resemblance to the god Frey rather suggests that he was regarded as the typical representative of a Golden Age in the past. The story of HeSinn and Hogni is very difficult to locate, both in regard to time and scene. Widsith however, which is our earliest authority for it, represents ^ The Origin of the English Nation, p. 134 f. 2 But also with Theodric. This is a question to which we shall have to refer later. 3 Venimus in ilium locum ubi dudum Vidigoia, Gothorum fortissimus, Sarmatuin dole occubuit (cf. also cap. 5). 28 THE HEROIC AGE [CHAP. Hagena (Hogni) as king of the Holmryge, who appear to have dwelt in eastern Pomerania. Since the whole of the south coast of the Baltic had probably become Slavonic by the end of the fifth century, it is at all events unlikely that the story refers to any period after this. We shall have to discuss later how far these stories are to be regarded as historical and to what extent the characters and incidents with which they deal are to be attributed to myth or fiction. Here it is sufficient to point out that with the exception of the story of HeSinn and Hogni, the connections of which are obscure, all the stories which we have just been discussing are referred by our authorities to generations anterior to the characters of Beowulf Hence we may safely conclude that the period embraced by the common poetry and traditions of the various Teutonic peoples — what we may call the Heroic Age of these peoples — had come to an end by the middle of the sixth century or at least by the death of Alboin. Its upper limit must in view of the evidence given above be set from two to three centuries back — probably three centuries if we include the story of Ostrogotha. These limitations are clearly such as to call for some attempt at explanation. Why do the cycles of story which are common to the various Teutonic peoples mention no historical character later than Alboin ? Before we can hope to give a satisfactory answer to this question, a number of other phenomena will have to be taken into account. One or two observations however will not be out of place here. The period extending backwards from two to three centuries before the reign of Alboin coincides with what is generally known as the Age of National Migrations (Vblkerwanderungs- zeit). It was during this period that many of the Teutonic nations broke through the frontiers of the Roman empire and carved out for themselves extensive kingdoms within its terri- tories. Among these were the realms of Guthhere and Theodric, and in part also that of Attila. There is no doubt that in all these cases the conquest of the Roman provinces brought with it a great accession of wealth and profoundly affected the life of the invaders. II] THE HEROIC AGE 29 The same period witnessed the conversion of most of the continental Teutonic peoples to Christianity, another change which produced far-reaching effects upon them. Yet it is not clear at first sight how this change is connected with the chrono- logical limitation of the stories, for while some of the chief characters, Attila for instance and doubtless Eormenric, were heathens, others such as Theodric were certainly Christians. The change of faith is not a motive which plays any part in the stories themselves. Whatever weight we may be disposed to attach to these observations, it should be noted that they do not seem to apply to every case. Thus we shall see in the course of the next chapter that Danish characters figure more prominently than those of any other nation, not only in Scandinavian but also in English records, throughout the period ending with Hrolfr Kraki. Yet the Danes took no part, collectively at least, in the movements against the Roman Empire, nor did Christianity penetrate to them before the ninth century. It is worth remark- ing therefore that stories relating to Denmark stop where they do, and that for centuries after the time of Hrolfr we can scarcely give the name of a single Danish prince. In the following chapters it will be convenient to speak of the period which we have been discussing simply as the Heroic Age. The term 'heroic poetry,' as a translation of Helden- dichtung or Heltedigtning, may of course be applied in a sense to such works as Hdkonarmdl or the poem on the battle of Maldon, just as well as to Beowulf or the Hildebrandslied. But no ambiguity will arise if we limit the term 'heroic' here to what may be called the * Teutonic ' Heroic Age (das germa- nische Heldenalter), i.e. to the period embraced by the common poetry and traditions of the various Teutonic peoples. CHAPTER III. SCENE AND NATIONALITY IN THE HEROIC STORIES. In the last chapter it was pointed out that the age covered by the heroic poetry and traditions of the Teutonic peoples coincides with a clearly marked period of history, extending over about two or possibly three hundred years, and coming to an end in the latter half of the sixth century. Something must now be said regarding the geographical and ethnographical limitations of the stories — the localities in which the scenes are laid and the nationalities to which the various characters be- longed. The scenes are distributed over a considerable part of Europe, extending from Italy to Sweden and from western Russia to the Vosges and the Netherlands. The British Isles however seem to have lain outside the area, though in the late form in which some of the stories have come down to us, we do occasionally find references to them — generally to Scotland or Ireland — which are probably due to confusion with stories of the Viking Age. Indeed it is remarkable that the early Anglo- Saxon poems contain no reference to persons or events con- nected with this country. Further, except possibly in the case of the story of Hengest and Horsa, we have no evidence worth consideration that poems dealing with such subjects ever existed. Norway also is not made the scene of any of the main stories, though it is mentioned incidentally in English and German as well as Scandinavian poems. The Balkan peninsula figures only in the later German poems, while references to places in Italy are limited practically to the Dietrich (Theodric) and Wolfdietrich cycles. CHAP. Ill] SCENE AND NATIONALITY 3 1 Turning to the question of nationality we find the fol- lowing peoples represented : (i) to the Goths belong Eormenric, Theodric, Wudga (Witege) and probably most of the heroes associated with them ; (ii) to the Huns Attila ; (iii) to the Burgundians Guthhere and his family ; (iv) to the Rugii ap- parently Hagena (Hogni), the father of Hild ; (v) to the Franks Hugdietrich and Wolfdietrich, if the identifications are correct ; (vi) to the Frisians Finn ; (vii) to the Angli Wermund and Offa; (viii) to the Danes (a) Fr6?5i the Peaceful, (d) Sigarr and his family, together probably with HagbarSr and Haki, (c) Hrolfr Kraki and his family, perhaps also Froda (Frotho IV) and Ingeld; (ix) to the Gotar Hygelac, Beowulf and their relatives; (x) to the Swedes (Svear) ASils and his family. It is somewhat remarkable that we have no stories dealing with Alamannic or Bavarian heroes, since the German poems which have come down to us are almost entirely derived from the territories of these peoples^ The Vandals too are unrepre- sented, and probably also the Visigoths, while the evidence for Frankish heroes is slight and rather unsatisfactory, Frankish nationality is claimed by most scholars for Siegfried, chiefly on the ground that Xanten is represented as his home in the Nibelungenlied^. Yet he is never called a Frank, and it is not clear that Xanten was in the possession of the Franks at the time to which the story refers. The same nationality may perhaps be claimed for Waldhere. In several German poems Langres is said to be his home, though he is also called a Spaniard, while Ekkehard makes his father king of Aquitaine=*. To this question we shall have to return later. In any case both stories refer to a period considerably anterior to the real conquest of Gaul by the Franks, and it is certainly a curious fact that Clovis and his great achievements seem to be entirely unnoticed in poetry. It appears then that though most of the principal Teutonic nations are represented in our stories the relative prominence ' These poems do introduce Bavarian characters, such as the Markgraf Riidiger ; but they are not found elsewhere in heroic poetry. 2 In the prose piece Frd daW^i Sinfiotla (in the Older Edda) Sigmundr is said to have held territories in the land of the Franks. * Hence many scholars regard him as a Visigoth. 32 SCENE AND NATIONALITY [CHAP. assigned to them does not at all correspond to what we should expect. Most remarkable is the fact that in stories relating to the Continent nearly all the chief characters (Eormenric, Theo- dric, Guthhere, Attila, etc.) belong to nations which had passed out of existence before the end of the sixth century. From Jordanes (cap. 5) we gather that the preservation of the early Gothic traditions was very largely due to the pride taken by that people in its own heroes of the past. Yet it is not easy to see how the survival of the stories which have come down to us can be ascribed to any such feeling. Let us first examine the Anglo-Saxon poems. In Beowulf the scene is laid first in the land of the Danes and later in that of the Gotar. The hero himself belongs to the latter nation, but in the earlier part of the poem the former are decidedly the more prominent. Taking it as a whole the interest is divided between the royal families of these two nations ; the only other dynasty which comes in for any considerable share of attention is that of the Swedes. In the story of Finn the interest is centred in a prince and his followers who according to Beowulf were of Danish nationality and involved in hostilities with the Frisians. In Waldhere the hero and heroine, whatever their nationality, belong to Gaul, while their opponents are Burgun- dians. In Deor the interests are mainly, perhaps exclusively, Gothic. In Widsith the foremost characters are Eadgils, prince of the Myrgingas (a dynasty hostile to the Angli), and the Gothic king Eormenric ; after them the English king Offa, the Danish kings Hrothgar and Hrothwulf, Guthhere, king of the Burgundians, Aelfwine (Alboin, king of the Langobardi) and several Gothic heroes. In all these poems there is no reference, as far as we know, to any person of English nationality except Qffa and his relatives (Beow. 1944 ff.), nor except in Widsith is the name of the Angli even mentioned. If we turn now to the Scandinavian records, which are entirely Norse (Norwegian-Icelandic), so far as the vernacular literature is concerned, the phenomena which confront us are on the whole very similar. As we might expect from the com- parative lateness of our authorities the nationality of the various characters is not very clearly indicated. It is remembered that Ill] SCENE AND NATIONALITY 33 Atli (Attila) belonged to the Huns and lormunrekr (Eormenric) to the Goths, but Gunnarr (Guthhere) is only once called lord of the Burgundians. As for SigurSr, his later adventures are uniformly located in the Rhineland, but the history of his family is generally connected with Denmark, which is also the scene of Helgi's exploits. Most noteworthy however is the fact that Norway is only mentioned once in the Older Edda, namely in the prose of HelgakviSa HiorvartSssonar. In the sagas it is somewhat more prominent, e.g. in the account of Biarki's origin in Hrolfs Saga Kraka ; but these passages are usually regarded as accretions to the original stories. Of course there are numerous other sagas which deal exclusively with Norwegian history and legend. These stories however are peculiar to Norwegian- Icelandic literature, and the earliest persons who figure in them, if we may regard them as historical, cannot have lived before the .seventh century. In stories relating to earlier times the scene is practically always laid in Denmark or southern Sweden or in the lands south of the Baltic. Even in the German poems national feeling has influenced the choice of subjects comparatively little. The poems in their present form are mainly Austrian or Bavarian. Yet except in the second half of the Nibelungenlied this region does not figure prominently^ and even there the Bavarian characters that occur are generally believed to be rather late additions to the story. The chief characters of the story in its original form were clearly Burgundians and Huns, to whom the Goths may have been added at a fairly early date. The stories which deal with the Rhineland — those of Siegfried and Walther — may be derived ultimately from early Prankish poems; but this cannot be proved. Most of the others are concerned with Gothic heroes, the true scene of whose adventures is to be sought in Poland, Hungary, Italy and other countries which had ceased to be Teutonic before the time of our authorities. Theodric, it is true, seems to have become a national hero in the south-east, but this feature is prominent only in the latest poems, and even here it is clearly remembered that he belonged to Italy. But the most 1 Tyrol is the scene of several of Dietrich's and Wolfdietrich's adventures ; but generally it is the southern (Italian) part of that country. 34 SCENE AND NATIONALITY [CHAP. remarkable case is that of Kudriin. The poem itself is probably Austrian, but the names which it contains show that the story is derived from Frisian sources. The scene is laid first in Ireland, then in Denmark and lastly in Normandy. Yet if we take into account the various Scandinavian versions and the references to the story in Anglo-Saxon poetry, there is every probability that it came originally from the Baltic. This short discussion will suffice to show how singularly free the poems we have been discussing are from anything in the nature of national interest or sentiment. They are certainly national in the sense that the characters are drawn entirely, or almost entirely, from within the Teutonic world — for even Attila can hardly be regarded as an exception. But nationalism in the narrower sense, i.e. in the interests of the poet's own nation or tribe, seems to be altogether wanting. The interest is centred in one or more individual characters and in the various adven- tures that befall them. Sometimes, as in Beowulf, it does also embrace the history of the family to which these persons belonged, but the nation, apart from the royal family, is practically disregarded. The contrast afforded by the historical poems of the ninth and tenth centuries^ is sufficiently obvious. We have seen that these poems, whether English or German, uniformly deal with the poet's own nation. The poem on the battle of Brunanburh is an expression of national triumph. It is not concerned with the personal adventures of the king or his brother, but with the prowess of the English army as a whole. The bravery of the princes is certainly noticed, but they appear to be regarded as the champions and representatives of the nation. The Lud- wigslied breathes on the whole a similar spirit, in spite of its strongly religious tone. Even in the skaldic poetry of the North traces of national pride are clearly discernible, as in Hdkonarmdl (v. 3), where Haakon, at the head of his Norwegian troops, is described as the terror of the Danes. It may perhaps be urged that, though the poems which have come down to us have no national interest, they may be derived ^ The poem on the battle of Maldon approximates much more closely than any other of this age to the spirit of the old poetry. Ill] SCENE AND NATIONALITY 35 from older poems which originated in the hero's own land. Thus many scholars believe Beowulf to be of Scandinavian origin in one sense or another, though the linguistic arguments which have been brought forward in favour of this view are not generally admitted. But there is a curious lack of uniformity in the national interests of the poem, as we have already seen. If it had been the chief intention of the original poet to glorify the Danish nation, he would not have ignored it as he has done in the latter part of the poem. On the other hand if his inten- tion was to glorify the Gotar he would hardly have begun with an account of the early kings of the Danes. The difficulty has been got over by supposing that the poem as we have it is of composite formation, and it may very well be that the second part of the poem is a later addition. But it is to be observed that somewhat similar phenomena occur in other cases. Thus the stories of Siegfried and Attila are connected both in the Nibelungenlied and the Edda, and there can be no doubt that the connection is of considerable antiquity. Yet the only common element in the two stories is supplied by the Burgun- dians, and the portraiture of their princes, especially that of the king, is hardly of such a character as to suggest its derivation from a poem composed for the glorification of the Burgundian nation. The evidence then of Beowulf alone is scarcely sufficient to justify us in assuming more than that its author or authors were interested in the royal families of the North, and that they possessed a considerable amount of information regarding them. The account of the early kings of the Danes seems to be in the nature of a tribal or family tradition — to be compared with the early stories given by Jordanes, Paulus Diaconus and Widukind. Traditions of this kind are no doubt generally of a mythical character, and consequently their origin is to be sought in the particular locality or family with which they are concerned. We have no evidence that such traditions formed the main theme of stories which were common to the poetry of the various Teutonic peoples. But it is natural enough that a poet who was well acquainted with some royal family, whether that of his own 36 SCENE AND NATIONALITY [CHAP. nation or not, would also know its traditions, and that he would utilise them incidentally or by way of introduction in a poem largely concerned with the fortunes of that family. The main story of the poem stands on a different footing. Of course if it could be shown that the Danish princes who figure both in Beowulf and the various Scandinavian records were fictitious persons, who never really existed, we should be bound to hold that they were derived from a common story, probably of Scandinavian origin. But few scholars would now be willing to admit such a proposition. Certain incidents, such as the exhortation of Ingeld (Ingellus) by the old warrior (Starcatherus), may be held to point to a common origin in poem or saga ; but most of the events narrated appear to have been either preserved by memory or invented independently. With the stories of SigurSr (Siegfried) and Waldhere the case is somewhat different. It is the opinion of the great majority of scholars that both these heroes are mythical or fictitious, in spite of the fact that they are associated with undoubtedly historical characters. If this view is correct — a question which we shall have to discuss later — we may conclude at once that the different versions of the two stories, Scandinavian and German in the one case, English and German in the other, have sprung from a common source, whether in poem or saga. But even if we take the opposite view, viz. that SigurtJr and Waldhere were real persons and that their adventures are founded on fact, it does not by any means follow that the different versions of their stories must have originated independently. Neither hero seems to have belonged to a family of outstanding position, nor were their exploits such as to influence the destiny of nations^ In the age of Hunnish supremacy scores of petty princes must have undergone somewhat similar adventures and distinguished themselves by similar deeds of heroism. Hence it can hardly be due to accident that the handful whose names we know were celebrated far and wide in the Teutonic world. ^ The victories over the Saxons and Danes described in the Nibelungenlied and the Thattr af Nornagesti have little in common and are scarcely to be regarded as an essential feature in the story of SiguriSr. Ill] SCENE AND NATIONALITY 37 The story of Eormenric is again rather a different case. There can be no question that this king was a historical person, but the earliest detailed account which we possess of his doings, viz. that given by Jordanes, dates from nearly two centuries after his death. Now we find what is perhaps the most striking episode in Eormenric's story, namely the account of Swanhild and the vengeance attempted by her brothers, both in Jordanes and the Older Edda (GuSrunarhvot and HamSismdl) with comparatively slight variations. The chief feature wanting in Jordanes' ac- count is supplied by the Annals of Quedlinburg^ How much truth the story contains we are not in a position to decide. But even if we grant that the main features are historical, the event can hardly have been of the first importance, since the attack seems really to have failed in its object. Ammianus MarcelHnus says that Eormenric committed suicide owing to despair at the impending Hunnish invasion, and Jordanes recognises that his death was partly due to this cause. But in the later (Scandi- navian and German) accounts it is entirely forgotten. Hence if we bear in mind the close agreement between the Gothic and Scandinavian versions of the story, not only in the names of the characters but also in the description of Swanhild's death, we can hardly help inferring that they are derived from a common narrative source. In conclusion mention may be made of a story which appears to be definitely at variance with historical truth, namely the account of Attila's death given in the Edda poems AtlakviSa and Atlamdl. It is there stated that Attila was murdered with his two children by his wife Gu?5run in revenge for her brothers (Gunnarr and Hogni), whom he had treacherously put to death. Now Jordanes (cap. 49) says that Attila died from the bursting of a blood-vessel on the night of his marriage with a girl named Ildico. As his account is derived from Priscus, a contemporary and trustworthy writer, there can be little doubt that it is correct. Yet it should be observed that the Roman chronicler Marcellinus Comes, who wrote apparently a few years before Jordanes, says * Anastasius annos XXVII Ermanrici regis Gothorum a fratribns Hemido et Serilo et Adaccaro, quorum patrem. interfecerat, amputatis manibus et pedibus turpiter, lit dicmus erat. occisio (Mon. Germ.. Scr. ill n. ■i\\. rtlo et Aaaccaro, quorum patrem interjecerat, am^ dignus erat, occisio (Mon. Germ., Scr. in p. 31) 38 SCENE AND NATIONALITY [CHAP. that Attila died by the hand of a woman\ The account given in the Edda therefore is no invention of an Icelandic or Norwegian poet, but founded on a story which was current among the Romans within a century after Attila's death. Indeed considering the circumstances it is by no means unlikely that the story originated immediately after the event. These examples will be sufficient to show that the subjects of many of our poems are derived from stories which passed from one Teutonic people to another and some of which were of great antiquity. Further, it is a proof of the popularity of these stories that they were preserved until comparatively late in the Middle Ages, in spite of the fact that they did not appeal to national interests. The question how they were preserved and transmitted is one which we shall have to discuss in the following chapters. We may note at once however that the most obvious means of preservation, namely by means of writing, was almost certainly not used to any great extent. Had that been the case the divergencies between the different versions of the stories would be far less noticeable than they are. Even in the case of Waldhere, which shows probably the least amount of variation, the Anglo-Saxon fragments show a treatment of the subject totally different from that which appears in Ekkehard's poem. Again, there is no evidence that the Roman alphabet was used in the North, except possibly by a few foreigners here and there, before the end of the tenth century, while the Runic alphabet, though it had been known for many centuries, seems not to have been employed for literary purposes until very late times. But, as we have already mentioned, some of the Edda poems date probably from the ninth century, and the story of HeSinn and Hogni is used by the poet Bragi who lived apparently in the early part of that century. On the Continent of course the case is somewhat different. We have seen that Charlemagne did have a number of ancient poems written down ; but there is nothing to show that his collection had any permanent influence. When the Quedlinburg annalist or his authority quotes the ' Ind. vn. Actio et Studio coss. (a.u. 454) : Attila rex Hunnorum Europae orbator prottinciae noctu mulieris manu cultroque confoditur. quidam uero sanguinis reiectione necatum perhibent. Ill] SCENE AND NATIONALITY 39 heroic stories, perhaps some two centuries later, he refers not to any written works but to songs formerly current among the country people. It is doubtless by oral tradition therefore, whether in verse or prose, that the stories of the Heroic Age have mainly been preserved. There can be no question that a large number of similar stories have perished. A glance through the catalogues of Widsith will show many names which otherwise are entirely unknown to us, and also an appreciable number which are not mentioned elsewhere in Anglo-Saxon literature, though they figure in German and Scandinavian records. It would scarcely be wise however to assume that all the stories of the Heroic Age were common Teutonic property. Thus we have no evidence for the story of Waldhere in the Norths while stories dealing with Danish heroes seem to have been little known in Germany. The latter observation deserves notice all the more in view of the obvious popularity of such stories in England — a fact proved not merely by Beowulf and Finn, but still more by the prominence assigned to Danish characters in Widsith. It will be convenient now to summarise briefly the results of our discussion. We have seen that the scenes of the heroic stories are distributed over most of the lands formerly occupied by the Teutonic peoples. Norway and England however, as well as the distant kingdoms of the Visigoths and Vandals, seem to lie outside the area. The heroes also are drawn from many nations, though not in the proportion which we should expect. Thus the Prankish nation, which ultimately became dominant, is but poorly represented, while the most prominent places are taken by peoples such as the Ostrogoths and Burgundians, which lost their nationality in the course of the sixth century. We have further seen that a very large proportion of the characters of the heroic stories figure in the literature of two or more nations, and that frequently the same stories are told of them. In the latter case it is probable, indeed often practically certain, that the different versions of the story are ultimately derived from a common narrative. On the other hand, where the same characters are known but only in connection with ^ Excluding of course ThitJreks S. af Bern. 40 SCENE AND NATIONALITY [CHAP. Ill different events, such derivation can be proved only if it can be shown that the characters themselves are fictitious. This remark applies especially to a number of characters common to Beowulf and Scandinavian stories relating to Hr61fr Kraki and his times. It is fully in accord with these facts that the heroic poems are not concerned at all — or at least only to a very slight degree — with local or tribal interests. Their tone indeed may be described as in a sense international, though with the restriction that characters and scenes alike are drawn exclusively from within the Teutonic world. CHAPTER IV. THE ORIGIN AND HISTORY OF THE HEROIC POEMS. In an earlier chapter (p. 3) it was mentioned that the Enghsh heroic poems are usually ascribed to the seventh or eighth centuries. We must now try to see whether any means are to be found of dating their composition more precisely. Unfortunately very few references to the poems or their subjects occur in works which can be dated with anything like certainty. The most important is contained in a letter from Alcuin to Hygebald, bishop of Lindisfarne, written in the year 797 : " When priests dine together let the words of God be read. It is fitting on such occasions to listen to a reader, not to a harpist, to the discourses of the fathers, not to the poems of the heathen. What has Ingeld to do with Christ.? Strait is the house ; it will not be able to hold them both. The king of heaven will have no part with so-called kings who are heathen and damned ; for the one king reigns eternally in heaven, the other, the heathen, is damned and groans in hell. In your houses the voices of those who read should be heard, not a rabble of those who make merry in the streets ^" From this passage it is clear that at the end of the eighth century there were current in Northumbria certain poems, probably well known poems, dealing with a heathen king named Ingeld, whom we need not ^ Verba Dei legantur in sacerdotali conuiido. ibi decet ledorem audiri, non citharistam ; sermones patrum, non carmina gentilium. quid Hinieldus cum Christo? angusta est domus ; utrosque tenere non poterit. non uult rex coelestis cum paganis et perditis nominetenus regibus communionem habere, quia rex ille aelernus regnat in coelo, ille pagantis perditus plangit in inferno, uoces legentium audiri in domibus tuis, non ridentium turuam in plateis. Mon. Germ., Epist. Carol. II 124; cf. O. Janicke, ZfdA. XV 314. 42 THE ORIGIN OF THE POEMS [CHAP. hesitate to identify with Ingeld the son of Froda, who figures in Beowulf Of course it is not at all likely that the reference is to Beowulf itself, for the part played by Ingeld in that poem is insignificant. Acquaintance with the subjects of the heroic poems is shown also by a mistake in the Historia Brittonum, 31, which dates probably from about the same period. This passage contains a genealogy, tracing the descent of Hengest and Horsa from Woden and of Woden from Geat. The latter part is known also from many other texts, in which it regularly runs as follows: Woden Frealafing, Frealaf FrWmwidfing, Frithuwidf Fmning, Finn Godwulfing, Godwidf Geatmg. In the Historia Brittonum however in place of Finn Godwidfing we find Finn qui fnit Folcwald — which is clearly due to confusion with Finn the son of Folcwalda {Finn Folcwalding), a Frisian king mentioned in Beowulf and Widsith, as well as in the fragment which bears his name. Further evidence is afforded by names of persons and places. There can be no doubt that even in the seventh century it was customary to take the names of famous men of the past or present. Danihel, bishop of Winchester (d. 745), and lohannes, bishop of Hexham (d. 721), are instances which no one will dispute. In the Durham Liber Vitae we meet with the names Aethan and Cundigeorn. It must not be assumed that persons bearing such names were necessarily of Celtic blood. Indeed the spelling suggests rather that they were Englishmen called after Aidan and Kentigern. Deusdedit, archbishop of Canter- bury (655 — 664), doubtless derived his name from Pope Deus- dedit (615—618), while the West Saxon king Ceadwalla (685 — 688) was in all probability called after the British king of the same name, who died in 642. It is extremely likely that Hlothhere, king of Kent (673 — 685), obtained his name from one of the Frankish kings, Lothair II (584 — 628) or Lothair III (656 — 670), for the element hloth- is not used elsewhere in Anglo-Saxon names. Even in the sixth century we hear of English princes who seem to be called after Frankish or Gothic kings of the same period. Thus Tytla, the name of the father of the East Anglian king Redwald, is probably taken from the IV] THE ORIGIN OF THE POEMS 43 Gothic king Totila ; it is not of an English type. Two sons of the Northumbrian king Ida were called Theodric and Aethelric, perhaps after the Gothic king Theodric and his successor Athalaric. As the element theod- is somewhat rare in England, it is not unlikely that the Northumbrian prince Theodbald, a son of Aethelric, derived his name from the Frankish king Theodbald. The occurrence of such names as Widsith and Beowulf {Biutilf) in the Liber Vitae shows that names were taken not only from contemporary persons and from books but also from native poems and traditions. Indeed researches which have been made in this direction have demonstrated that names of the latter type were extremely popular. But it has not been sufficiently pointed out that such names occur most frequently in the earliest times and gradually become more rare — a fact which is of considerable importance for our purpose. The total number of personal names found in the five poems Beowulf, Finn, Waldhere, Widsith and Deor is .I32\ and of these altogether 57 recur as names of persons mentioned in English historical documents. Over forty of these names belonged to persons who appear to have lived, or at any rate to have been born, before the end of the seventh century^ while at least thirteen of them are unknown after the same period. To the latter class belong the important names Widsith and Beowulf In local nomenclature it is possible to trace at least 51 of the 132 names mentioned above. In some cases these names may have been taken direct from the story, e.g. when we find in Kent two localities close together called Hokes clif and Hengstes earns (Birch, Cart. Sax., Ill 1000). A similar case, very frequently cited ^ is that of Beowanhammes hecgan and Grendelcs mere in Wiltshire {ib. II 6'] J'), though neither of these names is included in our list. But in the majority of cases it is more probable that the place-name is taken in the first instance from that of a previous landowner, and consequently that the connection with ^ The names Grendel, Cain, Abel, Alexandreas and Casere are not included. "^ For the figures and the method of calculation see Note I. 3 But open to very serious objections, as has been shown by Prof. W. W. Lawrence in the Publications of the Modem Language Association of America, xxiv 251 ff. 44 THE ORIGIN OF THE POEMS [CHAP. the Story is only secondary. Hence it is important to notice that out of the 51 place-names no less than 19 contain names which are not borne by persons in historical documents. The explanation of this lies doubtless in the fact that the place-names for the most part became fixed at a very early period, and consequently that they exhibit an earlier stratum of personal nomenclature. If we add the place-names to the personal names the total number of heroic names found in England in historical documents seems to be ']6. Out of this number only seven apparently are limited to persons born after the end of the seventh century, and of these again almost all occur in the course of the eighth century. These statistics show clearly that such names were most popular during the sixth and seventh centuries, especially if we bear in mind that the materials for this period are incomparably less than those for the following three centuries. Hence, if we are justified in drawing any conclusions from nomenclature, the popularity of the heroic stories was distinctly on the wane in Alcuin's time. The argument from nomenclature holds good of course only for showing the popularity of the stories ; it cannot prove the existence of the poems which we now possess. In one case however we may probably make an exception. The name Widsith is obviously fictitious' and based on the travels with which the minstrel is credited. The introduction, in which alone the name occurs, is in all probability a later composition than the rest of the poem^ and designed to explain what follows. It is of importance therefore to note that, if we may judge from the place in which this name occurs in the Liber Vitae, it must have been borne by a person of the seventh century. ^ Compound names containing wtd- or -s't\ {-sin^-) are used in other Teutonic languages; but the latter apparently does not occur in England, while the former is extremely rare. 2 In contrast with the body of the poem (w. lo — 134) it is non-strophic, after the general fashion of Anglo-Saxon poetry. Originally when the poem was recited it may have been introduced with a short explanation in prose, such as we find e.g. in Rfgsmal or Atlakvi'Sa. The epilogue (vv. 135—143), which is likewise non-strophic, may belong to the same stratum. IV] THE ORIGIN OF THE POEMS 45 We must next turn to the internal evidence. The linguistic criteria are of a somewhat unsatisfactory nature and investiga- tions in this field have led to few definite results. It is clear that the heroic poems do not exhibit any dialect in its purity — a remark which is true of Anglo-Saxon poetry in general. West Saxon forms predominate and there is no doubt that the final recension of the text is due to scribes who employed this dialect. Yet at the same time there are a sufficient number of Midland or Northumbrian characteristics^ to render it highly probable that the poems were not only composed but also originally written down in one of these dialects. Beyond this however no safe conclusions can be attained owing to the lateness of the MSS. In regard to syntax the heroic poems are at least as archaic as any other remains of Anglo-Saxon poetry which have come down to us. We may notice especially the use of the definite article, which in reality is still a demonstrative pronoun in the heroic poems. It occurs comparatively seldom in connection with a weak adjective followed by a substantive — a usage which is nearly universal in most of the Christian poems. Thus in Cynewulf's works the proportion of examples with and without the article varies from 7:1 to 9 : i, and even in the first part of Guthlac, which is believed to date from soon after the middle of the eighth century, the proportion is 7 : i '^ ; but in Beowulf it is only I : 5. If the Dream of the Cross, in which the proportion is 2 : I, is rightly attributed to the early years of the eighth century ^ it seems reasonable to date the composition of Beowulf quite half a century further back. The nearest approach to the usage of Beowulf is shown by Exodus, in which the proportion is over 2 : 3. Unfortunately we have no certain means of dating this work, though it is generally believed to be one of the earliest of the Christian poems. Its archaic character would be natural enough if it is really the work of Caedmon, who flourished while ^ Especially the regular use of unsyncopated forms such as 3 sg. on-winded, past part, onsended. * The statistics for these poems are as follows : Juliana 27 : 3, Christ (II) 28 : 3, Elene 66 : 9, Guthlac (A) 42 : 6, Dream of the Cross 10 : 5, Exodus 10 : 14, Beowulf 13 : 65 ; see Brandl, S.B. d. Akad. der Wiss. zu Berlin, 1905, p. 718 f. 3 Cf. Brandl, I.e. (p. 721 ff.). 46 THE ORIGIN OF THE POEMS [CHAP. Hild was abbess at Whitby (658 — 680) and who according to Bede^ did compose a poem or poems on this subject. The metrical characteristics of the heroic poems differ but little from those of Anglo-Saxon poetry in general. Cases of absence of contraction after the loss of intervocalic -h- (e.g. in the half-verse hean hiises) can be paralleled in poems dating from the close of the eighth century or even later, where they are doubt- less to be regarded as poetic archaisms. On the other hand importance is generally attached to the absence of any evidence for the retention of -u after a long syllable and to the shortening of syllables containing -r- which was originally followed by antevocalic -h-. Thus it is contended that such combinations as to widan feore in the latter half of a verse cannot go back to the middle of the seventh century ^ since the form in use at that time would be feorha, which would offend against the metre. As a matter of fact half-verses of the condemned type do occur in Anglo-Saxon poetry, Beowulf itself containing at least eight examples. But even if we were to admit all these statements and emend the offending verses the argument would be con- clusive only on the assumption that the poems were written down from the very beginning^ Poems which are preserved by oral tradition alone are manifestly liable to small verbal changes, especially in a metre so flexible as that of the Teutonic alliterative verse. Thus in place of the expression to widan feore we find occasionally widan feore in the same sense, and even in Beowulf we meet with widan feorh which is not improbably the oldest form of the phrase. Before the loss of final -u it would be a perfectly regular half-verse, but the operation of this change would render it impossible and necessitate the substitution of a synonymous expression. In principle, it should be observed, the ^ Hist. Eccl., IV 24 : canebat autem de creatione mundi et ortgine humani generis et tota genesis histo7-ia, de egressu Israel ex Aegypto et ingressu in t err am repromissionis, de aliis phirimis sacrae scripturae historiis, de incarnatione dominica, passione, resurrectione et ascensione in caelum, de Spiritus Sancti adiientu et apostolorum doctrina. item de terrore futuri iudicii et horrore poenae gehennalis ac dulceditie regni caelestis multa carmina faciebat ; sed et perplura de beneficiis et iudiciis diuinis, etc. Prof. Brandl {^Grundr., \\ 1028) holds that the reference is to lyric poems throughout. But is this interpretation really necessary ? 2 On this date see Note II. * For a brief discussion of this question see Note III. IV] THE ORIGIN OF THE POEMS 47 assumption of such substitutions seems to be absolutely necessary, unless we are prepared to deny that any old poems or even verses survived the period of apocope. Yet there is a sufficient amount of resemblance between English and German poetry, not merely in the general metrical scheme but also in the construc- tion of individual verses^ to render such a conclusion extremely improbable. Consequently I am very much inclined to doubt whether any safe conclusions as to the date of the poems can be obtained from metrical considerations, except of course as regards their final form. Of far greater importance is the fact that with the exception of the Finn-fragment, which consists of only fifty verses, all our poems contain passages or references of a religious (Christian) character. In Beowulf alone there are about seventy such passages of which the significance is not open to question, and seven or eight others which may belong to the same category. Out of the total number thirty-three are limited to single verses or half-verses^, while another sixteen affect not more than two verses in each easel The longest passage of all (v. 1724 ff.) contains at least 37 verses, the next longest (v. 175 ff.) fourteen. The rest vary from three to nine verses*. The theology which appears in these passages is of a singularly vague type. There are four distinct references to incidents in the early part of Genesis, viz. one (v. 90 ff.) to the Creation, two (vv. 107 ff., 1 261 ff.) to the story of Cain and Abel and one (v. 1688 ff.) to the Flood. Apart from these there appears to be no reference to any passage in the Bible except perhaps in v. 1745 ff., which are thought by some to be based on Ephes. vi. 16, and in v. 3069, which contains the phrase ' day of judgment.' We find ^ A few examples are given in the following chapter. 2 vv. 27, 72, loi, 570, 670, 706, 711, 756 (?), 786, 788, 790, 801, 806, 811, 852, 940, 967, 975, 986 (?), 1201, 1255, 1379, 1626, 1658, 1680, 1682, 2088, 2182, 2216 (?), 2276 (?), 2469, 2650, 3083 (?). » vv. 168 f. (?), 227 f., 440 f., 478 f., 588 f., 625 f., 945 f., 955 f., 1314 f., 1397 f., 1778 f., 1841 f., 1997 f., 2819 f., 2874 f., 3108 f. * w. 13—17, 90—8, 106—14, 316—8, 381—4, 665—7 (?), 685—7, 696—702, 928—31, 977—9, 1056—62, 1261—5, 1271— 6, 1553—6, 1609— 11, 1661—4, 1688 — 93, 1716— 8, 2291—3, 2329-31, 2341— 3 (?), 2741—3, 2794—7, 2855—9, 3054—7, 3069—73. 48 THE ORIGIN OF THE POEMS [CHAP. also a few references to rewards and punishments in a future life^ The word god is of very frequent occurrence and always used in the Christian sense. The other epithets of the Deity are ' lord ' {frea, dryhten), ' father ' {faeder), ' creator ' {scyppend), ' ruler ' {waldend), ' almighty ' {alwalda, aelmihtigd), ' ruler of men ' {ylda or fira waldend), ' ruler of glory ' {wuldres waldend)^ 'shepherd of glory ' {wiildres /tyrde),'k'mg of glory' (wu/durcyuing), 'guider of the heavens' (rodera raedend), 'helm of the heavens' (Jieofena helm), ' ruler of victories ' {sigora waldend), ' king of victories ' {sigora so'^cyning). On the other hand there is no example of the word gast in a religious sense (Holy Ghost), nor of the name Crist, nor of any epithet denoting 'Saviour' {jiergend, haelend etc.). Hardly less curious is the total absence of the word engel, for expressions such as ' lord of angels ' {engla dryhten"^) are among the most frequent epithets of the Deity in Anglo-Saxon religious poems. Lastly, there are no references to the saints, to the cross or to the church, nor to any Christian rites or ceremonies. It appears then that the religious utterances of the poem are of a singularly one-sided character. Indeed it has been observed^ that, with the exception perhaps of vv. 977 — 9, " their theology is covered by the Old Testament, and a pious Jew would have no difficulty in assenting to them all." Certainly the facts are such as to call for some explanation, especially since the religious poems are pervaded by a wholly different tone. One suggestion is that Beowulf was composed under the influence of the missionaries from lona ; but it is extremely doubtful whether the influence of Irish Christianity would tend in this direction at all*. Another is that the poet had little direct knowledge of the Christian religion, but that he was acquainted with some religious poems. This explanation certainly seems to fit the case much better than the other. Moreover there is one piece of positive evidence in its favour. 1 vv. 588 f., 977— 9> 2741—3. 2819 f. ^ It is perhaps worth noting that in v. ■2186 the expression dryhten wereda is used of Hygelac. Elsewhere in Anglo-Saxon poetry this phrase is applied only to the Deity. ^ Clark Hall, Beowulf, p. xxviii. * If there is Celtic influence at all it is more probably Welsh, IV] THE ORIGIN OF THE POEMS 49 In Beow. v. 89 ff. we hear of recitation to the accompaniment of the harp in Hrothgar's hall, and the subject of the recitation is the creation of the world. It appears to me highly probable that we have here an allusion to Caedmon's poem or poems on Genesis, which may very well have been among the earliest of that poet's productions. At all events it was by his hymn on the Creation that he first became known. The inference is strengthened by the rather close resemblance which the hymn bears to the phraseology of Beowulf. If the two poets were contemporary the author of Beowulf would have no other Christian poet on whom to draw, and the limitations of his theological equipment might be satisfactorily accounted for on the hypothesis that he knew only a few of Caedmon's works. As a matter of fact two or three out of the list given by Bede' would have been quite sufficient to provide him with all the statements and terms that he uses. There is another question however with regard to the composition of Beowulf which has aroused more controversy than this, namely whether the Christian passages formed an original part of the poem or not. In the former case of course the poem cannot have been composed before the second quarter of the seventh century. Indeed, if we grant the use of Caedmon's poetry the earliest possible date would be about 660. On the other hand if the Christian passages are due to inter- polation the upper limit for the dating of the poem vanishes into air. As to the possibility of such interpolation in principle we need scarcely entertain any doubt. It is true that the Christian passages or references cannot as a rule be removed without breaking into the rhythm. Consequently, if interpolation has taken place we must assume it to be the work of poets or minstrels, and not of scribes. But have we any reason for doubting that the minstrels of that period were capable of such ' interpolation.' Wherever poetry — at all events anonymous narrative poetry — is preserved exclusively by oral tradition, it is usually the case that the minstrel is allowed a certain amount ^ Cf. p. 46, note, where the passage is quoted in full. 50 THE ORIGIN OF THE POEMS [CHAP. of freedom in the presentation of his subject'. Now probably no one will suggest that it was only after their conversion to Christianity that the English began to compose poems about ' heathen kings.' But, if we grant that such poems were already in existence, does it really involve a greater amount of effort on the part of the minstrels to bring these poems up to date — by removing objectionable matter and introducing expressions in accordance with the new religion — than to compose an entirely new set of poems on the same subjects. I cannot think that such a view will be seriously maintained. Therefore we must consider the case of Beowulf — the only narrative poem which has come down to us entire — without prejudice on the general question ; and we must endeavour to see whether it bears the stamp of a new composition or that of an old work which has been brought into conformity with new ideas. The probability or improbability of the latter view will of course depend largely on the amount of inconsistency which the poem is found to contain. Until within the last few years the majority of scholars believed that Beowulf was a composite work. This theory was most fully developed in the writings of Miillenhoff and ten Brink. According to the former- the poem was made up from four separate lays, though in its present form nearly half of it is the work of interpolators. The latter^ likewise traced the origin of the poem to lays, but explained its inconsistencies as being due not to extensive interpolations but to the combination of two parallel versions. In regard to the relative antiquity of the various parts of the poem there was great divergence of opinion both between these scholars and generally. It is perhaps partly on this account that in recent years there has been a reaction in favour of believing that the poem as we have it is practically the work of one man, though it is allowed that he may have made use of earlier lays. But those who have 1 The amount of freedom diflfers of course greatly from case to case (cf. Note IV, p. loi ff.) ; but it is only in communities which have elaborated the art of minstrelsy to a very high degree that the form of words can become absolutely stereotyped. * Beowu// {i8Sg), pp. no — 160. ' Quellen unci Forschungen, LXii (1888) ; summarised p. 242 flf. IV] THE ORIGIN OF THE POEMS 5 1 adopted this view seem to agree that the author, whatever his precise date, belonged to the Christian period, and consequently that the religious passages are not due to interpolation. Now in the first place it is clear that the story of Beowulf is derived from the Baltic, and the first question which we have to settle is as to the time at which the information on which it is based became known in England. The Angli themselves were originally a Baltic people, as I have tried elsewhere to show, and there is no doubt that down to the time of the invasion of Britain they were thoroughly familiar with all the surrounding regions. But we have no evidence whatever for believing that such was the case within the historical period. By the end of the sixth century, when the first missionaries arrived in this country, they had apparently ceased to be a sea- faring people, and we have no record of any voyage made by an Englishman across the North Sea for several centuries. Again, the Danes became familiar to the west of Europe during the sixth century ; but from about 580 onwards we hear no more of their presence on the North Sea for fully two centuries. During the whole of this period their name is heard of only in connection with the missionary expeditions of St Willibrord, early in the eighth century. I have suggested elsewhere^ that their temporary disappearance was due to the maritime supremacy held by the Frisians. At all events we have archaeo- logical evidence for a considerable amount of communication between southern Norway and the Frisian coasts during this period, while for the Baltic such evidence is almost wholly wanting. Bearing these facts in mind we can hardly doubt that the information used by Beowulf was acquired before the end of the sixth century — in all probability we may say considerably before that date"''. Next we have to notice that we have practically no trustworthy information regarding the history of the English kingdoms before the middle of the sixth century, and I think it ^ The Origin of the English Nation, p. 93, note. ^ The references quoted on p. 41 ff. preclude the possibility that these stories were first acquired from the Danes, when the latter again became known in this country about the close of the eighth century. 52 THE ORIGIN OF THE POEMS [CHAP. will be the opinion of any attentive student of early English history that even the best informed persons of Bede's time were not much better off in this respect than we ourselves are. How then are we to account for the preservation of detailed infor- mation regarding the early kings of the Danes and Swedes ? The only answer to this question, so far as I can see, is that the doings of such persons must have become embodied in stories which were preserved by recitation in a more or less fixed form of words. Such recitative pieces may have consisted of poetry alone or of poetry mixed with prose, like some of the pieces contained in the Older Edda. If we may trust the analogy of what appear to be the oldest pieces in this collection, such as Volundarkvi?5a or HelgakviSa Hundingsbana II, the speeches would be given in metre, while the connecting narrative might be partly or wholly in prose and quite brief. We have no evidence for believing that the early Teutonic peoples ever used entirely prose narratives, like the Icelandic and Irish sagas, for such purposes. At all events it seems to me that if Beowulf is no older than the middle of the seventh century we are bound to assume the existence of earlier poems or narratives on the same subject. Such pieces may of course have been quite short, and it is likely enough that our epic has made use of more than one of them. One perhaps may have dealt with the hero's exploits at the Danish court and another with his last adventure, while in the scene between Beowulf and Hygelac it is possible that an older poem has been incorporated, more or less complete, in the text^ But we have yet to take account of what is perhaps the most striking feature of the poem, namely the fact that, though it abounds in expressions of Christian sentiment, yet the customs and ceremonies to which it alludes are uniformly heathen. Among these we may mention the funeral ship in V. 27 ff., the offerings at the shrines in v. 175 f., the observation of the omens in v. 204 and the curious reference to hanging in V. 2444 ff. {c{. V. 2939 ff.), probably also the use of the boar on ^ A different view is taken by Schucking, Beowulfs Riickkehr (Studien zur engl. Philologie, XXl), p. 65 ff. IV] THE ORIGIN OF THE POEMS 53 helmets (vv. 303 f, 11 1 1 f., 1286, I45iff., 2152) and the burial of the treasure (v. 2233 ff.), together with the curse imprecated on the person who should disturb it (v. 3069 ff.). But most important of all are the descriptions of the disposal of the dead by cremation in vv. 1 108 ff., 2124 ff., 3137 ff. In the long account of Beowulf's obsequies — beginning with the dying king's in- junction (v. 2802 ff.) to construct for him a lofty barrow on the edge of the cliff, and ending with the scene of the twelve princes riding round the barrow, proclaiming the dead man's exploits — we have the most detailed description of an early Teutonic funeral which has come down to us, and one of which the accuracy is confirmed in every point by archaeological or contemporary literary evidenced Such an account must have been composed within living memory of a time when ceremonies of this kind were still actually in use. The significance of these passages seems to me to have been altogether misapprehended by recent writers. If the poem preserves its original form and is the work of a Christian, it is difificult to see why the poet should go out of his way in v. 175 ff. to represent the Danes as offering heathen sacrifices; for not long before he has introduced a song of the Creation at the Danish court, and in the sequel Hrothgar is constantly giving utterance to Christian sentiments. Again why should he lay Beowulf himself to rest with heathen obsequies, described in all possible detail, when in his dying speeches (vv. 2739 ff., 2794 ff.^) the hero has been made to express his faith and gratitude to the Almighty ? On the other hand if the poem was originally a heathen work these inconsistencies are perfectly natural. If it was to retain its place after the change of faith and to be recited in the presence of bishops or clergy, all references to actual heathen worship or belief would of necessity have to be either accompanied by censure — as is the case in the homiletic verses following v. 175 ff. — or else suppressed altogether, and their place taken by expressions in accordance ^ We may refer especially to the account of Attila's funeral given by Jordanes, cap. 49 (from Priscus). ^ Apart from certain expressions the general tone of these speeches, especially the last words of all (v. 2813 ff.), is scarcely Christian ; but they contain nothing which is obviously opposed to Christian doctrine. 54 THE ORIGIN OF THE POEMS [CHAP. with Christian doctrine. Hence it seems to me probable that such expressions are frequently in the nature of substitutions for objectionable matter, rather than gratuitous additions ; and in the same way I would account for the occasional survival of ideas which appear to be essentially heathen S though they are cloaked in Christian phraseology. But references to practices such as cremation which, though heathen, had long ago passed out of use, would not excite the same repugnance and conse- quently might be allowed to stand. It may be urged" that cremation seems to have lingered on among the Old Saxons of the Continent until late in the eighth century. True : but it is quite incredible that a Christian poet should borrow from this quarter a method of funeral for his Christianised heroes. If the description of Beowulf's obsequies stood alone a bare possibility might be conceded to the sugges- tion that it had once formed a poem by itself, unconnected with Beowulf, and based upon a traveller's story. But cremation is clearly regarded as the normal rite throughout the poem, apart from the legendary story of Scyld. We have another description of it in the episode dealing with Finn (v. i io8 ff.), and above all there is the purely incidental reference in v. 2124 fif. : " Yet when morning came the knights of the Danes could not burn his (Aeschere's) lifeless form with fire, nor lay the man they loved on the pyre. She had carried the body away," etc. Here the poet realises the significance^ of the rite quite clearly and consequently notes that the inability of the Danes to carry it out added materially to their sorrow. In such a case the possibility of Christian authorship seems to me to be definitely excluded. On the hypothesis that these descriptions had come down from the days of English heathenism all is easily explicable. At the time when the poem was Christianised it may very well not have been known that the rite of cremation was still practised among the heathen of the Continent, and in later ^ E.g. in the imprecation, v. 3069 ff. The imprecatory formulae of charters can scarcely be regarded as analogous. ^ Cf. Brandl, op. cit., p. 1003. ' The same idea is frequently expressed in the Homeric poems, e.g. II. vn 79 f., XXII 342 f., XXIII 75 f , XXIV 37 f., Od. XI 71 ff., etc. IV] THE ORIGIN OF THE POEMS 55 days the verses of the old poet would be handed on in parrot fashion without their significance being generally understood. Well informed persons however, like Alcuin, who had travelled abroad, perceived clearly enough that, however much coated over with Christian phraseology, the heroic poems were in reality of an essentially heathen character. Now cremation was widely prevalent in this country during the early days of the Saxon invasion — a fact attested by numerous cemeteries especially in the northern and midland counties, including the valley of the Thames. But it appears to have become a thing of the past when the Roman missionaries arrived here ; otherwise it is difficult to account for the absence of any reference to the custom in the records which have come down to us. Indeed we may say with safety that it had passed out of general use, at least in the southern half of England, quite a generation before this time ; for there are scarcely any traces of it to be found in those western districts which appear to have been conquered during the latter half of the sixth century. Consequently, if we are justified in believing that the descriptions of cremation ceremonies contained in Beowulf date from a time when the practice was still remembered, we must conclude that they were composed not later than the third or fourth decade of the seventh century ^ But it is not contended, so far as I am aware, by any scholar that the account of Beowulfs obsequies belongs to the earlier parts of the poem. It is the final scene of the story, it is not contained in any speech, and further it is of a thoroughly epic character and would be quite out of place in a short lay. Hence, if the line of argument which we have been following is legitimate, we shall be forced to admit that though the poem has undergone a fairly thorough revision in early Christian times, it must in the main have been in existence some time before the conversion. I do not mean to suggest that the ' revision ' was entirely limited to the religious element. Other changes and additions may have been made about the same * This date does not depend in any way on the question where the poem originated. Cremation may possibly have lingered in Northumbria longer than elsewhere ; but that kingdom seems to have become entirely Christian between 626 and 642. 56 THE ORIGIN OF THE POEMS [CHAP. time^ What I do mean is that the great bulk of the poem must have been in existence — not merely as a collection of lays or stories, but in full epic form — an appreciable time before the middle of the seventh century. The other heroic poems do not furnish us with any similar criteria for estimating the date of their composition, but there seems to be no valid reason for doubting that they are quite as early. Two of them, Deor and Widsith, are expressed in the first person and lay claim to being of a remote antiquity. Deor says that he had been the bard of the Heodeningas and that he had been displaced by a skilful minstrel named Heorrenda. Since in old Norse literature Hia^ningar (i.e. Heodeningas) means ' HeSinn and his men,"^ and since Heorrenda can hardly be separated from the minstrel Horand in Kudrun, it would seem that the poet claims to have been a contemporary of HeSinn and Hogni, with whose story we have dealt briefly above (pp. 8 f , i6). Widsith is still more explicit. The poet states that he visited the Gothic king Eormenric, who as we know died about 370. It is true that incidentally he mentions that he had met with a number of other princes, some of whom lived in the fifth and sixth centuries ; but the visit to Eormenric is his main theme. Eormenric is of course one of the most prominent figures of the Heroic Age, and it may be for this reason — as the type of a powerful king — that he is chosen for the poet's host and patron. But then it is by no means so easy to see why he is associated with such an obscure person as Eadgils, prince of the Myrgingas. The suggestion that the poem is founded upon a tradition that this Eadgils possessed a famous minstrel breaks down upon the name Widsith, which is obviously fictitious as we have seen (p. 44). It appears to me that considerably less difficulty is involved in the hypothesis that the kernel of the poem* is really the work of an unknown bard ^ E.g. possibly some of the elegiac passages (e.g. vv. ■2236 — 2270, 2450 — 2464), which show a certain resemblance to such poems as the Ruin and the Wanderer. 2 The name Hegelinge in Kudrfin is probably a corruption of Hetelinge (i.e. Heodeningas). •' Presumably including vv. 88 — 108; but I am not prepared to suggest an elaborate analysis of the poem. IV] THE ORIGIN OF THE POEMS 57 of the fourth century, and that successive minstrels from time to time have added the names of famous heroes with which they were acquainted^ — a process to which the original plan of the poem may well have offered inducement. However this may be neither of the poems shows any characteristics which suggest a later date than Beowulf Both appear to be constructed in strophic form, a feature rare else- where in Anglo-Saxon poetry, while Deor also has a refrain, which is almost without parallel. Lastly, we have seen that Widsith occurs as a personal name, apparently in the seventh century, and that this presupposes the existence not only of the poem itself but also of the introduction, which is clearly later. The case of Waldhere stands somewhat apart from the others, since it has been suggested that this is really a transla- tion of a lost German poem, on which Ekkehard's Waltharius is also based. That the story came from the Continent may of course be granted ; but we have also to consider when and in what form it was brought. The linguistic arguments which have been adduced in favour of the German origin of the poem are not now generally maintained. But it is further to be noticed that the poet seems to have treated his subject very differently from Ekkehard. The speeches, with which the fragments are entirely taken up, have nothing corresponding to them in the Latin poem, while the characterisation of the heroine is as unlike as it well could be. Ekkehard represents her as a timid creature, but in the fragments she displays a spirit which may fairly be called martial. It is unwise to lay stress on agreements between Ekkehard and the fragments as against the version of the story given in ThiSreks Saga af Bern. The less complicated form of the latter — in which Guthhere is omitted and Hagena represented as an officer of Attila, pursuing the fugitives — may be due either to imperfect acquaintance with the story or, perhaps more probably, to the conditions under which it had been preserved. We shall see later that for a ^ These lists are perhaps derived in part from mnemonic catalogues — ' inventories ' of the stories known to the minstrels who composed them. Metrical catalogues of this kind are said to be in use among Servian minstrels at the present day ; cf. Krauss, Slavische Volkforschungen, p. i86 ff., where a specimen is given. 58 THE ORIGIN OF THE POEMS [CHAP. considerable period heroic poetry appears to have been entirely neglected by the higher classes in Germany ; and it may be accepted as generally true that when stories are preserved only by the peasants complex situations tend to become simplified, while all except the most prominent characters drop out. As for the date at which the story became known in England we may note that besides Aetla and Hagena, which may come from other sources, Waldhere, Hildegyth and Hereric were all names current during the seventh century. There seems therefore to be no adequate reason for believing Waldhere to have had a different history from the other heroic poems. On the whole, taking all the poems, including Beowulf, together, we may conclude with probability that they assumed substantially their present form^ in the course of the seventh century. But if our reasoning with regard to the composition of Beowulf is correct we shall have to refer the first treatment of the subject to the sixth century, i.e. almost if not quite to the Heroic Age itself Deor and Widsith may quite possibly contain still older elements. We may now turn to the Old Norse poems. Here the data at our disposal are of a very different character, for the metrical evidence is said to preclude the possibility that any of the extant poems date from before the ninth century. It may perhaps be questioned whether all of them are necessarily new compositions since that time — whether certain of them may not be old poems somewhat recast. To this question however we can hardly hope to obtain a satisfactory answer. The fragmentary Ragnarsdrapa of Bragi Boddason, who seems to have lived in the first half of the ninth century, is probably the earliest extant piece which refers to stories of the Heroic Age. In this poem we find allusions both to the story of HetSinn and Hogni and to the attack made upon lormunrekr by HamSir and Sorli. Thiodolfr's Ynglingatal, perhaps half a century later, contains references to the story of HagbarSr and ^ Waldhere, Deor and Widsith all contain ' Christian ' passages, like Beowulf. The interpolations in Widsith (e.g. vv. 15 f., 82 ff.) appear to have been made by some one who possessed a certain amount of erudition ; but there is no need to attribute them to a different period. IV] THE ORIGIN OF THE POEMS 59 Signy, as well as brief accounts of the Swedish kings Ottarr and A8ils, who are mentioned in Beowulf (cf. p. 20). All these stories, except that of Ottarr, are told also by Saxo, but in a somewhat different form, which points to their derivation from Danish rather than Norse sources'. On the other hand the stories of the Volsungar, Sigmundr and SigurSr, are not mentioned by any early Danish authority. The story of Sigur?5r is generally supposed to have been intro- duced into Norway from Germany ; and in some sense or other this would seem necessarily to be true, since in the Northern version, as well as in the German, the scene is laid chiefly in the Rhineland. But it is apparently impossible to determine when and in what form the story was transmitted. We have already noticed that there is archaeological evidence for a considerable amount of communication between Norway (not Denmark) and the southern (Frisian) coasts of the North Sea during the seventh and eighth centuries, and this is clearly a factor which deserves to be taken into account. Further, it is worth noting that, with the exception perhaps of Atli"-, the names all appear in regular Northern form', as if they had been known from the earliest times, e.g. Gunnarr, Hogni, Giiiki, Bu^li. This consideration, as far as it goes, certainly favours a very early date ; but it is hardly conclusive^ The story of Sigmundr stands on a somewhat different footing. In the first place Sigemund (Sigmundr) himself is little more than a name in German tradition, while though ^ Cf. Olrik, Kildeme til Sakses Oldhistorie, p. 132. 2 The name appears to have been quite common in the North during the Viking Age ; yet the apparent absence of umlaut suggests derivation from a (Frisian ?) form corresponding to the Ang.-Sax. Aetla (cf. Aecci, Aeddi beside Acca, Adda). * In contrast (e.g.) with Kftdr^tn, which clearly shows its foreign origin (cf. p. 34). * Such names may have been current before, though their frequent occurrence is no proof of this and may be due to the popularity of the heroic poems. It is perhaps worth noting that alliteration is shown by certain names which are generally believed to have been introduced into the story in Norway or Iceland, e.g. Oddriin, Erpr, EUill with Atli, Giajlaug, Gullrbnd (Gu'Srtln?) with Gtuki and Gunnarr. If we may judge from the genealogies in Landnamabok and elsewhere the principle of alliteration seems to have been generally given up in family names before the ninth century. 6o THE ORIGIN OF THE POEMS [CHAP. Welsung and Sintarfizzilo occur as personal names, they are not connected in any way with the story of Siegfried. Again, in the Helgi poems, which contain no reference to SigurSr, Sigmundr is connected with the Baltic, and this is still more clearly the case with his son Helgi, who is unknown to the German story. Thirdly, in Beowulf, which knows Sigemund and Fitela, though not SigurSr, the former is brought into juxtaposition, and apparently also into comparison, with a Danish prince named Heremod. The same two persons are brought together also in the Old Norse poem HyndlulioS, while the Hakonarmdl likewise seems to imply some connection between them, as I have tried to show elsewhere^ Apart from the passages specified this Herm6t5r (Heremod) is apparently not mentioned in Scandinavian literature, but the facts noted seem to indicate that the two characters were connected in poetry before English and Danish tradition became separated, i.e. presumably in the sixth century. We have seen that many of the persons mentioned in the main narrative of Beowulf were remembered also in Scandinavian tradition. But since these persons are in all probability to be regarded as historical, it is hardly safe to infer the existence of ancient Scandinavian poems, unless the same incidents are related of them, which is generally not the case. There is a rather striking verbal resemblance^ between the first speech of Wiglaf (Beow. 2633 — 60) and certain passages in Biarkamal (especially Saxo's version), where Hialti is addressing Biarki ; and this fact is the more noteworthy if Biarki is really to be identified with Beowulf. But the words themselves are of a somewhat general character and might have been used on other occasions. Again there is a certain afifinity^ between the account of the dragon-fight in Beowulf and that of a similar incident related of Frotho I by Saxo (p. 38 f.) ; and here again a connection can be traced indirectly between the two heroes. But the story of Frotho seems really rather to resemble the account given in Beowulf of Sigemund's dragon-fight ; so it 1 The Cult of Othin, p. 5 1 f. 2 Cf. Bugge, Beitrdge, xii 45 ff. ^ Cf. Sievers, Ges. d. Wiss. zu Leipzig, Ber. 1895, p. 180 if. IV] THE ORIGIN OF THE POEMS 6l may be questioned whether the points of affinity between the two did not originally form part of a standard description of incidents of this kind. A clearer case is that of the poems attributed by Saxo to Starcatherus, in which that warrior exhorts Ingellus to avenge his father. These are clearly to be connected with the speech of the old warrior to Ingeld in Beowulf (vv. 2047 — 2056), though there is little verbal resemblance. Moreover we have seen (p. 41 f.) that in Alcuin's time poems dealing with Ingeld were known and probably popular in England. The relationship of the passage in Beowulf to these may be compared with that of another passage (vv. 1068 — 1159) to the fragment dealing with Finn. The poems on Ingeld given by Saxo are traced by Prof Olrik^ to a Danish source; and there can be little doubt that his view is correct, as they share the characteristics exhibited by other stories which appear to come from the same quarter (cf p. in). Thus the queen's name is not given and her brothers are described simply as sons of Suertingus. Further, the story is cut right away from the surroundings in which we find it in Beowulf, and it may be that for a time it survived in Denmark only in ballad form. Yet, however much change it had undergone before it came under Saxo's treatment, there can be little doubt, in view of the English evidence, that its origin is ultimately to be sought in heroic poetry, or at all events heroic narrative, dating from the sixth century. Lastly we must mention the story of Uffo's single combat, though, strictly speaking, this is probably not of Danish origin. It was certainly well known in England and there is good reason for believing that its home is to be found in the district to which it refers, i.e. the neighbourhood of Angel, Slesvig and Rendsburg. I have tried elsewhere* to show that this story also rests on historical foundations. But the details of the combat, as given by Saxo and Svend Aagesen, and certain legendary features, such as the dumbness or silence of the hero, which are present in both the Danish and English versions of the story, strongly favour the view that it was embodied in ^ Kilderne ttl Sakses Oldhistorie, pp. i8 fif., 132. 2 The Origin of the English Nation, p. 118 ff. 62 THE ORIGIN OF THE POEMS [CHAP. poetic form at a very early period. On the other hand there is nothing to show that such poems survived till Saxo's time. The story is apparently unknown to all Norse authorities. Many of the German poems which have come down to us are known to be derived, directly or indirectly, from earlier ones, but regarding the antiquity of the latter nothing can be stated with certainty. The Hildebrandslied, which is the only extant piece of early poetry, goes back at all events to the eighth century. Further, the language^ used by Einhard in describing the poems collected by Charlemagne (cf. p. 5 f.) would scarcely be appropriate unless they were believed to be more than a century old by that time. We may probably therefore refer them at least to the seventh century. It seems likely that some of the lost poems of the Langobardi were of still greater antiquity. In the poem which celebrated their victory over the Vandals (cf p. 10), a story with which we shall have to deal more fully in a later chapter, a very prominent part appears to have been played by the heathen gods. Such a piece can hardly have been composed after the end of the fifth century, at which time the Langobardi were already Christians. Regarding the antiquity of Gothic heroic poetry there can be no question, for Jordanes, our chief authority on this subject, wrote about the middle of the sixth century, i.e. during the Heroic Age itself. We have already noticed (p. 37) that his account of Eormenric appears to be coloured by poetic tradition. But of the heroes whom he enumerates (cap. 5) as celebrated in poetry, the only one of whom we know anything, Vidigoia, is described as Gothornni fortissimus in a quotation (cap. 34) from Priscus, who lived about a century earlier^ There is good reason therefore for believing that the Goths possessed heroic poems as early as the first half of the fifth century. ^ Item barbara et antiquissima carmina, quibiis ueterum regum actus et bella canebantur, scripsit tnemoriaeque mandauit (Vita Caroli Magni, cap. •29). * Cassiodorus (Var. VIII 9) states that the Gothic king Gensimundus, who according to Jordanes (cap. 48) reigned shortly after Eormenric's death, was widely celebrated in poetry (toto orbe cantabilis) ; but his name is preserved only in Gothic records. IV] THE ORIGIN OF THE POEMS 63 We may now sum up briefly the results of our discussion. The heroic poetry of the Goths certainly belonged to the Heroic Age itself, and it is more than likely that certain Langobardic poems were nearly as old. Some heroic poems belonging to other German peoples may probably be referred at least to the seventh century. The chief monument of English heroic poetry must be ascribed to the first half of that century, while some of the other poems claim to be of greater antiquity. The lost heroic poetry of the Danes seems to have been occupied largely with the same subjects as the English poems, and since the stories generally refer to the Baltic we may reasonably infer that heroic poetry flourished in that region during the sixth century. On the whole then it seems probable that the develop- ment of heroic poetry began in the Heroic Age itself, not only among the Goths but throughout the greater part of the Teutonic world. 64 HEROIC NAMES IN ENGLAND [CHAP. NOTE I. THE USE OF HEROIC NAMES IN ENGLAND. The distribution of heroic names in English historical documents is as follows. The Liber Vitae contains 35 {^7), of which nine (eleven) are peculiar to that work^ Of those which occur in other documents 29 ^ belong to persons born apparently before 700, six to such persons only, 14 or 15^ to persons born between 700 and 800, two to such persons only, and 22* to persons born after 800, four to such persons only. It must be remembered that, apart from the Liber Vitae, the materials for the ninth and tenth centuries are much more abundant than those for earlier times. In the Liber Vitae itself it is possible within certain limits to distinguish between persons of early and later date. Investigations in the lists of kings, queens and abbots have shown that they are arranged chronologically, and it is only reasonable therefore to expect that the same is true of the much longer lists of clerics and monks. A brief examination of the names will make it clear that such is actually the case. Some modernisms of language, such as -/erl/i for -frith, or of orthography, such as -ht- for -ct-, are practically confined to the last parts of the lists, which may be by later hands^ But even within the parts certainly written by the first scribe archaic forms such as -iu- for -io- are more frequent at the beginning of the lists than later ; so also occasional -b- for -f-. Above all we may note the uncompounded names in -/, a type common in early Anglo-Saxon, but practically extinct after the eighth century. Fol. 24, the first in the clerics' list, contains eleven such names, while fol. 30 contains none ; fol. 34, the first in the monks' list, has eight, while fol. 38 has only one. Now we shall probably be well within the mark if we assume that one fifth of the names in each list^ belong to persons born in the seventh 1 37 (eleven), if we admit the emendation of the corrupt forms Vychgaz.x\A Vurmeri to Wydiga (Widia, Wudga) and Vyrmheri (Wyrmhere). See the Addenda. 2 Ecglaf, Eadgils, Eomaer, *Eormenric, Frod(a), Hereric, *Hoc, *Hrothmund, Ingeld, Offa, Oslaf, Sigemund, *Aehha, Sigeferth, Aetla, *Hagena, Theodric, Waldhere, Becca, Witta, VVada, Oswine, Sigehere, *Sceafthere, Alewih, Aelfwine, Eadwine, Wulfhere, Frithuric ; perhaps also Herebald. The asterisk denotes names limited to persons of the sixth and seventh centuries. 3 Eadgils, Eanmund, Ileardred, Hygelac, Ingeld, OfFa, Wermund, Weohstan, Wiglaf, Wada, *Scilling, Oswine, Sigehere, *Gislhere (perhaps Eomaer). ■• *Aelfhere, Ecglaf, Eanmund, Heremod, Offa, Sigemund, Wermund, Weohstan, Wiglaf, Wulfgar, Garwulf, *Ordlaf, Sigeferth, Waldhere, Becca, Wada, *Hun, *Hringwald, Aelfwine, Eadwine, Wulfhere, Frithuric (possibly also Deor). ^ Only the earlier lists, those printed in Sweet's Oldest English Texts, p. 154 if., are taken into account. 8 Lines 159 — 192 and 332 — 362 in Sweet's text. IV] HEROIC NAMES IN ENGLAND 65 century. On this basis we find that of the 35 (37) heroic names which occur in the Liber Vitae nine (ten) are limited to persons of the seventh century 1, and twelve (thirteen) to persons of the eighth 2. The total number of names in each century (cf. p. 43) may be obtained by comparing the evidence of the Liber Vitae with that of the other documents 3. The fact that so many heroic names occur in the Liber Vitae has led several writers to conclude that the true home of English heroic poetry was in the north. But no such inference is really justified by the evidence. The number of personal names recorded for the south of England during the sixth and seventh centuries is very small ; yet it is only here that we meet with persons called Eormenric, Hoc and Hagena — names which belong to quite different cycles of story. Moreover out of the 37 heroic names preserved in the Liber Vitae no less than 28 occur in place-names in various parts of England. The frequence therefore of heroic names in the Liber Vitae is to be attributed in part to the great abundance of the material and in part to its comparatively early date. Including the evidence of the Liber Vitae there are fifteen heroic names which seem to be borne only by persons of the eighth century or later times. Eight of these however occur also in place names*. Of the remaining seven ^ all except one or possibly two^ make their appearance during the eighth century. On the other hand it has already been mentioned that not less than nineteen heroic names'^ occur only in local nomenclature. We have still to mention a few heroic names which are not found in the extant remains of Anglo-Saxon poetry. Theodhere (Diether) is known only from the sixth century, Hild, Herding, luring, Omoling only from the 1 tBeowulf, tBilling, fFolcwald, Frod, Heremod, Hereric, Oslaf, fWidsith, Alewih (fWidia?). Names marked with a dagger are confined to the Liber Vitae. 2 Eadgils, tHildeburg, fHrothwulf, Wiglaf, Wulfgar, Garwulf, Aetla, Witta, tHeathuric, Sigeferth, Wulfhere, Frithuric (fWyrmhere?). The following names are found both in the early and late parts of the lists : Eanmund, fHama, Heardred, Hygelac, Ingeld, Oflfa, Sigemund, Wermund, Theodric, Wada, Aelfwine, Eadwine, fAegelmund ; perhaps also Herebald, if the abbot of this name is to be identified with the one mentioned by Bede {H. E. v 6). ^ For the details see the preceding notes. * Hildeburg, Hrothwulf, Wiglaf, Wulfgar, Ordlaf, Hun, Hringwald, Gislhere. ' Aelfhere, Weohstan, Garwulf, Scilling, Heathuric, Wyrmhere (?), Dior. ^ Aelfhere and Dior. The latter (in the form Diar) occurs only once (Birch, Cart. Sax. 497) and may be a mistake for Diara {ib. 507). The name Diora need not be of heroic origin ; it may be an abbreviation from such names as Diorwald, Diornoth. 7 Breca, Finn, Fitela, Hengest, Hnaef, Hrethel, Scyld, Weland, Guthhere, Geat, Gifeca, Heoden, Helm, Wald, Beaduca, Frithla, Secca, Gifeca — to which we may add Waelse in Walsingaham. Widia is also to be added, if it is not allowed for the Liber Vitae. For the list of place-names (not the personal names) I am dependent upon Binz, Beitr. xx 141 — 223. 66 HEROIC NAMES IN ENGLAND [CHAP. seventh, Hildegyth and Blaedla from the seventh and eighth, Wulf heard from all periods, Ecga only from the eighth century, Ecgheard from the eighth and ninth, Sigesteb only from the ninth. Hild, Wulfheard and Ecga are found also in place-names. In some of these cases, e.g. in that of Hildegyth, the non-occurrence of the name in the poems is clearly due to mere accident ; but it would scarcely be safe to assume that all these characters were celebrated in Anglo-Saxon poetry. NOTE II. ON THE DATING OF CERTAIN SOUND-CHANGES IN ANGLO-SAXON. In my Studies in Old English^, published in 1899, I endeavoured to formulate a scheme for dating approximately the chief sound-changes which took place in English during the first few centuries after the invasion of Britain. In the course of these investigations I was led to the following conclusions (pp. riy, 253 ff.) : i. that 'palatal umlaut' in Northumbrian and the dialect of the Vespasian Psalter took place before 650 ; ii. that the change from ae. \.o e (in all dialects except West Saxon) was in operation about 650 — 680 ; iii. that the loss of intersonantal h (in all dialects) belongs to the same period or a little later ; iv. that contraction through loss of intervocalic // may be dated roughly between 680—710; v. that the loss of final -u after long syllables and in words of the form - - ~- took place in all dialects at a time approximately contemporaneous with the operation of palatal umlaut in Northumbrian (i.e. before 650). Prof. Morsbach in his paper Zur Datierung des Beowulfepos'^ has dealt with several of the same problems and come to conclusions which differ widely from those at which I arrived. The chief differences are as follows : i. that after a long syllable which bore the chief accent -u was not lost before the end of the seventh century, though after a long ' nebentonig ' syllable the loss was somewhat earlier (p. 261 f ) ; ii. that intervocalic h was lost in Kentish by about 680, but in Mercian and Northumbrian the same change cannot be shown to have taken place before about 700 (p. 264) ; iii. that postconsonantal h (before vowels) was retained in Kentish in 679 ; its loss, at least in Mercian and Northumbrian, may be dated about 700, but after the loss of -u (p. 265). In summarising the results of his discussion (p. 273) he gives " about 700 " for the loss of postconsonantal h and " shortly before 700 " for the loss of -w. Incidentally he follows Biilbring {Elementarbuch, §§ 146, 528) in dating the origin of e (from 'West Germ.' a) before the breaking, and in placing the loss of h before / (in neolaecan) anterior to the operation of 'palatal umlaut' (monophthongisation). 1 Transactions of the Cambridge Philological Society, Vol. iv, Part II. The page- references are to the figures in the outer corners. 2 Nachrichten von der Konigl. Gesellsckaft der Wissenschaften zu Gbttingen, 1 906, pp. 251—277- IV] THE DATING OF CERTAIN SOUND-CHANGES 6/ The importance of this discussion for our present purpose^ Hes in the dates proposed for the loss of -u and of postconsonantal h. Prof. Morsbach concludes that Beowulf cannot have been composed before 700, since it contains a number of half- verses w^hich would have been metrically impossible before the operation of these changes, e.g. ofer fealone floditi), to widan feor{h)e. I have already expressed scepticism as to whether such inferences are really justifiable. This applies more particularly to the verses affected by the question of postconsonantal h, which are quite few in number. The date which Prof. Morsbach himself (p. 274) proposes for the composition of Beowulf is 700 — 730. He finds no difficulty in reconciling this with the statistics (given above, p. 45) for the use of the article. This seems to me rather strange ; yet the opinion of a scholar who stands in such deservedly high estimation cannot lightly be disregarded. Now let us examine the evidence on which these conclusions are based. First it will be convenient to take the loss of A. Prof. Morsbach holds that in Kentish intervocalic h was lost before postconsonantal h. The evidence is derived from a single charter issued by King Hlothhere in 679 (Birch, Cart. Sax. 45), which contains the place-name Vtiestan ae beside the personal name Velhisci (Latin Gen.). But surely conclusions of this kind are admissible only when a number of examples can be adduced. On the same principle we might argue from the name Irininredi (in the same charter) that the change se > e had taken place and also from the name Aedibnaeri (again in the same charter) that it had not. And what should we do with the earliest East Saxon charter (Birch, 81), in which the grantor is called both Oedelraedus and Ho{di)lredusl Again, it is clear that Bede wrote his own name Baeda ; but will anyone venture to hold that this represents the current pronunciation of his name in 731 — or indeed for some half a century earlier.'' In personal names we must clearly allow for traditional orthography. The form Irininredi may no doubt be used as evidence for the change ee. > t\ and similarly the form Vuestan ae may be used as evidence for the loss of h. But forms, especially personal names, like Velhisci zxvdi Aedilniaeri,\^\\\c\\ must long have been in use, may very well show an antiquated orthography — one which correctly represented the pronunciation of thirty or forty years previously. A single instance of such a kind is totally insufficient ground for supposing that the Kentish dialect treated postconsonantal and intervocalic h differently. Next we must consider the date given for the loss of h in Mercian and Northumbrian (p. 263 f ). I find some difficulty here in following Prof. Morsbach's line of argument. In Northumbrian there is, admittedly, no evidence at all for the preservation of h, while cases of its omission are numerous in Bede's History (written in 731), in addition to one or two instances in probably earlier authorities. For Mercian^ we are dependent ^ Prof. Morsbach's paper raises a number of questions besides those mentioned above. But I am obliged here to confine my attention to those which have a bearing on the dating of Beowulf. ^ The application of this term to the Epinal and Erfurt glossaries (or the arche- type) seems to me to be open to grave objection. 68 THE DATING OF CERTAIN SOUND-CHANGES [CHAP. on the Epinal, Erfurt and Corpus glossaries, the archetype of which is placed before 700 by Prof. Morsbach. In my Studies, p. 232, I came to the conclusion that in this archetype the cases of retention and omission of h were probably about equal in number. Prof. Morsbach replies that there is no necessity for such a conclusion, since all the extant glossaries themselves date from times when h was already lost. He himself decides^ against the loss of h in the archetype for two reasons : (i) because postconsonantal and intervocalic h are treated alike in the glossaries and the former was still retained in Kentish when the archetype was written 2; (2) because the assumption of such an early date for the loss of h would be incompatible with his own date for the loss of -u. The first of these arguments, it will be seen, rests upon the dating of the loss of postconsonantal h in Kentish, on which enough has been said above. The second depends upon a hypothesis which we shall have to consider presently. My reason for concluding that the loss of h occurred in the archetype was that in at least eight entries (probably several more) all three glossaries agree in showing forms without h. It is to be remembered that in these glossaries we are dealing not with independent documents but with copies made, more or less mechanically, from one original. This remark applies of course much more to Epinal and Erfurt than to Corpus ; for in the latter the materials have been rearranged, as well as augmented from other sources, while incidentally the forms have been modernised to a considerable extent. In Epinal such modernisation is not unknown, but it is restricted within very narrow limits, as may be seen from the use of ^ and/^. Further, it is to be remarked that we have no ground for assuming the language of the arche- type itself to have been consistent. The occurrence of numerous Dative forms and of expressions containing more than one word (e.g. per atiticipa- tionem — \orch obst) shows that the materials were derived largely from glosses in books, just as in the Leiden glossary*. Many of these glosses may have been written a generation or more earlier than the compilation of the archetype glossary. Whoever bears these facts in mind and at the same time compares the evidence for forms with and without h with that for/" and b will, I think, be forced to the conclusion that the forms with h do not represent the pronunciation of the compiler of the archetype, but that they were taken over by him from earlier sources. ^ If I have interpreted his meaning correctly. But I admit that I have had great difficulty in understanding the argument in paragraph 5 of p. 263 (especially the "mit -h-" of line 12). 2 Prof. Morsbach seems to regard Kentish as exceptional in its treatment of intervocalic k, rather than in that of postconsonantal k. I am not quite clear as to his reason for this. ' Cf. Studies in Old English, p. 240. * This text represents a more primitive type of glossary than the others and, though it is not an ancestor of theirs, it has without doubt used a considerable number of the same glosses (especially in Sweet's § XLv) which were incorporated in their archetype. IV] THE DATING OF CERTAIN SOUND-CHANGES 69 Now we may take the evidence for the proposed date of the loss of -ti. After enumerating (p. 253 fif.) a Hst of cases which have been suggested by various scholars Prof. Morsbach comes to the conclusion that the only certain example of a form in which -it is retained^ is the vjorA /lodu in the inscription on the Franks casket. As the whole theory largely depends on the value attached to this form we must consider it carefully. In my Studies, p. 156, I suggested that it should be regarded as an archaism and at the same time pointed out that -u is lost in another word {imneg) in the same inscription. Prof Morsbach rejects this explanation of unneg, which he connects, rightly, I think now, with O. Sax. nah. But this form can come perfectly well from *ndh{'w)u (earlier *nehwo), if not from *ndh'wa (cf. Goth. nehwa, nehw). That Ang.-Sax. neah, neh has lost an 2/-sound is, I contend, shown not only by the Gothic forms, but also by neolaecan and tieowest (cf. Lind. ^^wMwaJ), etc.). Prof Morsbach rejects my explanation of these forms also and adopts that of Prof. Biilbring, as stated above. But the latter is untenable ; for if e (from a) had come into existence before the operation of breaking, we should never find forms with ae, which as a matter of fact are fairly common in the earliest texts of all dialects. In particular we may note that the three extant coins of the Mercian king Aethelred (675 — 704) all have -raid. Hence the change ae, > e can hardly have taken place much before 650. Apart from iinneg"^, there are three forms on the recently discovered right side of the casket which may show loss of -u. Prof. Napier {An English Miscellatiy, p. 375 f.) is inclined to regard the forms sxr and daen (?), if not also /los, as Nom. sg. fem. But the interpretation of this part of the inscrip- tion is still uncertain in many details. I confess that since the discovery of the new side I am less inclined to regard the iormjiodu as an archaism than as a mere blunder. Even in the more intelligible parts of the inscription we find a number of forms which present serious difficulty : Romwalus, Reumivalus'^, gasric, Giu\easu. The last of these still seems to me to present the best illustration oi Jlodn. If the one is due to the loss of some letters — e.g. su for suiinsR) — the same may be the case with the other*. At all events an inscription which presents so many difficulties cannot be regarded as a safe authority on which to base a scheme for the chronology of sound-changes. The only other instance of -u on which Prof. Morsbach lays any weight ^ The form scanomodu on the solidus need not be taken into account. It is improbable that coins of this type were minted after the sixth century. - Prof. Morsbach further argues that even if my interpretation of unneg was correct it would prove nothing, since -u was probably lost after a long ' nebentonig ' syllable earlier than after a long ' haupttonig ' syllable. But neither the -gar of the Bewcastle inscription (cf. p. 70) nor the felt of the East Saxon charter can be admitted as evidence for this hypothesis. ' In spite of what is said by Prof. Morsbach (p. 264 f.) these forms are scarcely intelligible unless h was already lost. ■* E.g. perhaps flod u(p)aho/. •JO THE DATING OF CERTAIN SOUND-CHANGES [CHAP. is the form aetgaru in the Erfurt glossary. He speaks of it as 'nicht unwahrscheinlich ' (p. 257), though 'fraglich ' (p. 264). Many scholars cite it as an example of -« without reserve. Now in order to form a just estimate of the value of any form which occurs in the glossaries it is obviously necessary to take it in connection with the forms which the other texts show in the same entry. The entry in question {franiea — aetgaru) occurs in the Epinal and Corpus glossaries, as well as in that of Erfurt, though the two former have aetgaeru {setgaeni) for aetgaru. There can be no doubt that the relationship between the three glossaries is as follows : X (Arch. I) y (Arch. II) Epinal Erfurt Corpus though we do not know exactly how many intermediate stages lie between the extant texts and the original archetype. It will be seen that the question at issue is whether the Erfurt glossary or the other two have kept the original form — for if the archetype had contained both forms we may assume that some trace of the double entry would have been preserved (as in other cases). Now the Erfurt glossary is the latest of the three, it is the work of a foreign scribe and it is very carelessly copied. Moreover, no letter or combination of letters has suffered more than ae. Most usually this combination has been reduced to e ; but the loss of e is not infrequent, e.g. smal, hrad^ nacthegelae . In view of these facts it is unintelligible to me how anyone can uphold the evidence of the Erfurt glossary against the other two. But in this case it is used to prove the existence of an archaic form for which none of the glossaries elsewhere present a parallel. Lastly, we may remark that though it is frequently assumed that the word ^ar was an «-stem {*gaizu-) no evidence worth consideration has ever been adduced to prove it^. On all grounds therefore we are brought to the conclusion that the evidence for the preservation of -u in aetgaru is not merely open to question but entirely worthless. We must now notice certain early documents in which -7t is clearly lost. Prof Morsbach mentions the form felt in the earliest East Saxon charter (Birch, 81), which dates from 692 — 3. Here then -u was lost before the date in question^. But we can get back further than this, for the place mentioned is called Vuidnnindes felt. Some considerable time must have elapsed after the loss of -u before a noun, even a proper name, could change its inflection and adopt the endings of as was clearly still opera- tive. Will anyone suggest that this was the case after the middle of the seventh century .? Again, it can hardly be doubted that the form Osuiu., if not also Osuald, represents the pronunciation^ current in the time of that king. In Bede's History, ill 29, we have extracts from a Papal letter addressed to Osuiii (v. 1. Osuio) regi Saxoiium. The form os- can scarcely have become current in compounds before -u was lost in the Nom., Accus. sg. It is surely unnecessary to enter into further details. We have seen that ' Prof. Morsbach would assign the monument to the time of Aldfrith, who reigned 685 — 705 (not 725, a printer's error in Vietor's book). ^ In his own narrative of course he often uses forms which must have been antiquated in his time (e.g. Vurtigerno, Aedidni, Aeodbaldo) and also foreign forms in the names of persons of Continental origin (e.g. Agilberctus). But it may safely be assumed that all these cases are derived from earlier documents. ^ I.e. Oswiu (Oswald) ; cf. Baduuini, i.e. Baduwini. I have not taken account of the possibility that -uiu originally contained an -h-. If that could be proved the present discussion would be practically superfluous. 72 THE DATING OF CERTAIN SOUND-CHANGES [CHAP. practically the whole evidence for the proposed chronology consists of two forms, Vclhisci and Jlodu^ one of which is incapable of proving what it is meant to prove, while the other is of exceedingly doubtful value ; and on the other iiand that this chronology has opposed to it a large number of forms in the glossaries, in charters and in early Northumbrian authorities of various kinds. Now let us consider the various sound-changes in relation to one another. It is admitted, and necessarily so, that the loss of -ii took place before the loss of h. For the sake of convenience we may apply the terms ' Period I ' to times anterior to the loss of -u and ' Period II ' to the interval between the two changes. From neither of these periods have we any texts surviving, unless we are to reckon the inscription on the Franks casket to Period I. Our earliest charters, and apparently also the lost archetype of the glossaries, were composed at a later time ('Period 111') when // was no longer pro- nounced, though doubtless often written. But before the date of the earliest extant glossary (Epinal) a further change or changes had taken place which brought about the confusion of original t and f. It is clear from a com- parison of the glossaries that this confusion was later than the loss of h and also that it was almost, if not entirely, unknown to the archetype. So also in Northumbrian. In Bede's History we have no instance of ^ preserved in an English word, whereas examples of its omission are numerous. On the other hand it is clear that Bede usually retained the distinction between b (i.e. t) and / although the latter has already largely encroached on the former in the Moore MS.^ We may therefore constitute two subdivisions of the period subsequent to the loss of h, namely 'Period III' prior to the confusion of ^ and^J and 'Period IV' subsequent to this confusion. Now we can see clearly where to date the Franks casket, for it shows confusion of t and f in •ay/z/^-according to Prof Morsbach also in sefu. It belongs therefore not to Period I but to Period IV. Any attempt to fix an absolute chronology is of course rendered difficult by the absence of very early texts. We may probably assume that Period IV begins more or less about 700. The Kentish charter of 679 falls in Period 1 1 1. Now when -u was lost, h (whether intervocalic or postconsonantal) must have been a spirant — a fact which Prof. Morsbach seems to have entirely ignored. It is quite incredible that only a few years should elapse between that stage and the total loss of intervocalic h seen in the charter. The corresponding period in Germany lasted centuries, and I cannot conceive how the transition can have been accomplished anywhere in less than half a century. I con- clude therefore that the dates which I gave in my Studies were approximately correct. If there was an error it was in putting some of the changes slightly too late. I have little hesitation now in expressing my opinion that the loss of -u took place not later than the second or third decade of the seventh century. ^ Cf. my Studies, p. 247. It will be seen that the number of cases in which M, B and C agree in f (for 5) is very small. IV] LITERARY INFLUENCE IN BEOWULF 73 NOTE III. LITERARY INFLUENCE IN BEOWULF. I am not aware that any serious argument has been brought forward to show that Beowulf was a literary production. It is customary indeed, especially among English scholars, to use the word ' write ' with reference to its composition ; but this is frequently due to mere carelessness. Neverthe- less there is undoubtedly a widespread reluctance to admit that the poem came into existence without the use of writing — partly on account of its length and partly because its technique is of rather an advanced order. The first of these difficulties has now been definitely settled by the Servian poems, as we shall see in the next Note. Here we need concern ourselves only with the second. The most definite pronouncement on this subject known to me is contained in Prof Ker's book, The Dark Ages, p. 250. '■'■Beowulf and Waldere" he says, "are the work of educated men, and they were intended, no doubt, as books to read. They are not, like the Elder Edda, a collection of traditional oral poems." Here we have three distinct statements. That Beowulf and Waldhere are not a collection of poems, like the Edda, is manifest. Whether they are traditional oral poems at all is a different question, the answer to which depends in the first place on our attitude to Prof Ker's second statement — that they were intended as books to read. But this statement surely requires some evidence. The only argument brought forward is that "the Beowulf ^\s is intended as a book to be read, and is got up with some care. From the look of it, one places it naturally in the library of a great house or a monastic school ; and the contents of it have the same sort of association ; they do not belong to the unlearned in their present form." But, so far as the earlier part of this passage goes, the argument applies only to the tenth century. No one will deny that the earlier Anglo-Saxon poems were studied at that time ; indeed they had probably come to be regarded as, in a sense, classical. But have we any right to assume that scholars of the seventh or eighth centuries viewed these poems or their subjects in a similar light? If so how are we to account for the total absence of references to such subjects in the works of Bede? And what about Alcuin.'' Quid Hinieldus cu7n Christo? Is it likely that Alcuin would have regarded these pagani et perditi reges as suitable subjects for the attention of scholars.-' The natural presumption from his language is that he knew of them not from literary works, but only from street-minstrels whom he looked upon with disgust. Yet Bede and Alcuin can hardly have been ignorant of any important literary activity during their times, and it would be unfair to regard them merely as religious bigots. In view of their silence it seems to me a precarious hypothesis even that the poems were committed to writing much before the end of the eighth century. 74 LITERARY INFLUENCE IN BEOWULF [CHAP. There is no doubt of course that the writing of English was in common use during the early part of the ninth century. But the paucity of earlier evidence renders it probable that this was a recent innovation ; and the orthographical characteristics of eighth century documents point to the same conclusion. Laws were written in English from the beginning ; but we may safely assume that this was done by professional scribes, in all probability ecclesiastics. Otherwise there is little definite evidence for the writing of the vernacular, except of course in glosses. The more learned clergy clearly preferred to use Latin ; for the less learned Bede states {Ep. ad Ecgb.^ cap. 5) that he had himself had to make translations even of the Creed and the Lord's Prayer, for them to learn by heart. Where was a reading public to be looked for in such a period^.'' But we have yet to discuss the statement that " Beowulf and Waldere are the work of educated men." The question here is what is meant by the word 'educated.' Nowadays the expression 'educated Chinese' is used in more than one sense. Sometimes it is applied to those who have received a good education according to the traditional standard of that nation. Some- times however, especially in newspaper language, it is used only of those Chinese who have received a Western education. We need not doubt that the poets of Beowulf and Waldhere were among the best educated men of their day according to the traditional native standard. If they were court minstrels — a question we shall have to discuss in the next chapter — they could hardly be otherwise. But this is not what Prof. Ker means. The education which he has in view is of foreign (Roman) type, as may be seen by the latter part of the second of the quotations given above. Again (p. 252) he says : "The English epic is possibly due to Virgil and Statins ; possibly to Juvencus and other Christian poets, to the authors studied by Aldhelm and Bede." If so"'^, is it not remarkable that no obvious trace of such influence can be pointed out? It must not be assumed that the poets responsible for the composition or preservation of Beowulf would have any inclination to disguise their knowledge of foreign poetry. The use which they have made of incidents derived from the Bible is decisive evidence to the contrary, although their knowledge of this subject seems to have been of a most elementary description. In Widsith we actually have at least one ' The case of King Aldfrith shows that educated laymen were not entirely unknown ; but it is extremely unlikely that they were common. In one charter (Birch, 99) Wihtred, king of Kent, is made to say: signum sanctae crucis pro ignorantia litterarum expressi. This document exists only in a late copy ; but at all events it suggests that the practice of making the cross, instead of signing, was due to a wide- spread inability to write. ^ I confess that I am strongly inclined to suspect that anyone imbued with Latin learning would have lacked not only the inclination but also the ability to compose such a poem as Beowulf. This however is an opinion which could only be sub- stantiated by a wider knowledge of the history of poetry in various parts of the world than I can make any claim to possess. IV] LITERARY INFLUENCE IN BEOWULF 75 reference to a classical character (v. 15 ff.). But Beowulf is entirely free from anything of the kind. Prof BrandP has likewise been attracted by the idea that the growth of Anglo-Saxon epic poetry may have been due to Latin models. As a probable source of such influence he has fixed upon the Aeneid and even indicated a number of passages in this poem which may have suggested certain scenes and incidents in Beowulf Thus he notes that both poems begin with the story of a wanderer (Scyld, Aeneas) who came over the sea {feasceaft, primus et profugus) and founded a great dynasty or empire. Then he suggests a connection between the song of Hrothgar's minstrel on the Creation (Beow. 90 ff.) and a passage (Aen. i 742 fif.) which describes how Dido's minstrel lopas sang of the origin of men and beasts, among other cosmo- logical subjects. Further, he compares the whole of the scene which contains the latter passage with the account of Beowulf's reception at Heorot, noting the various stages in the arrival of the two heroes from their disembarkation to the feast with which they are welcomed in the palaces. Incidentally, he remarks that at the feast Wealhtheow hands the cup to Hrothgar and then to the visitors (Beow. 615 ff.), while Dido pays the same honour to Bitias (Aen. I 738). Lastly, he compares the racing after Grendel's overthrow (Beow. 864 ff.) with the rowing contest in memory of Anchises (Aen. v 104 ff.). I confess that, coming as they do towards the close of an admirable discussion of the subject, which no attentive student of Beowulf can fail to appreciate, the comparisons suggested by Prof Brandl strike me as sur- prising beyond measure. The resemblances between the athletic contests, so far as they have any existence at all, are due to practices which are world- wide. Is there any reason for supposing that the act of courtesy ascribed to Wealhtheow was not in full accordance with early Teutonic custom .'' Parallels may be found, easily enough, in Old Norse literature. The arrival of the wanderer (Scyld, not Beowulf) has nothing in common with that of Aeneas. The one is a baby and probably alone ; the other is the commander of a fleet. Then, as to the resemblance between Beowulf's visit to Hrothgar and Aeneas' visit to Dido, I can only say that I fail to detect its existence. Prof Brandl seems to lay most stress on this incident and points out certain parallelisms in the language : corripuere uiatn qua semita monstrat (Aen. I 418) with stig wisode gttmum aetgaedere (Beow. 320 f), and coram quern quaeritis adsum Trains Aeneas (Aen. I 595 f ) with Beowulf is min nama (Beow. 343). But are these not purely accidental coincidences, such as one could find between almost any two narrative poems ? When Prof Brandl speaks of Aeneas' " Verhandlung mit Dido zuerst durch eine etikettegemasse Mittelsperson," I do not understand what is meant. Certainly this description applies to the entry of Beowulf But Aeneas is present beforehand, shrouded in a cloud with which Venus has covered him. In the midst of the scene the cloud is suddenly parted and Aeneas disclosed to Dido's eyes. It seems to me that no meeting could well be more different. ^ Grundriss d. germ. Philol., II 1008. 76 LITERARY INFLUENCE IN BEOWULF [CHAP. IV There remains then only the fact that the two minstrels treat a somewhat similar theme (though in Beowulf this does not take place on the occasion of the hero's arrival). It is no doubt a curious coincidence ; but the introduction of such a subject in Beowulf may be accounted for quite satisfactorily without the hypothesis of any acquaintance with Virgil. I cannot quit this subject without remarking how much more plausible a case could be made out for deriving Beowulf frorn the Homeric poems, especially the Odyssey. Here there are striking parallels both in diction and terminology, as we shall see in a later chapter. We may note especially the epithets applied to princes and the formulae with which speeches are introduced. If we wish to find a real parallel to the reception of Beowulf by Hrothgar, it is provided by the account of Telemachos' visit to Menelaos. Similar parallels are to be obtained for the minstrel's lays and many other incidents. Are we then to suppose that Beowulf is based upon the Odyssey? That is a hypothesis which I will gladly leave for others to work out. For my own part I prefer the explanation that similar poetry is the outcome, or rather the expression, of similar social conditions. But in Beowulf and the Homeric poems, as against the Aeneid, we have additional common elements in the fact that the interest is centred in the actual characters — not in the destiny of their descendants — and in the vividness and reality of the narratives, in spite of Grendel, Scylla and the rest. The latter of these two common features seems to me to indicate that both the Greek and English poets were depicting types of life with which they were themselves familiar, whereas no one will dispute that the Aeneid is a product of the library. Haifa century ago, when the study of Teutonic antiquity was still young, there was a general eagerness to refer every institution and belief to a native origin. To-day we see the inevitable reaction — a hypercritical attitude towards every explanation of this character, coupled with a readiness to accept theories of biblical or classical influence on the slightest possible evidence. It is this intellectual atmosphere which, naturally enough, has given birth to the chimaera of a literary Beowulf — a creature which, if I am not mistaken, belongs to the same genus as certain well-known theories in Northern mythology. CHAPTER V. THE POETRY AND MINSTRELSY OF EARLY TIMES. In the preceding chapters we have seen that the persons and events celebrated in the heroic poems apparently all belonged to the fourth, fifth or sixth centuries, and further that heroic poetry was flourishing among the Goths during the same period. For the existence of English, Scandinavian or German heroic poetry at this time we have no absolutely conclusive evidence. But the materials from which our poems are formed must largely be referred to the sixth century. This may be seen most clearly in cases where the poems of two or more nations not merely treat an identical theme but also agree in the motif or in comparatively small details, as in the stories of Ingeld, Waldhere and Svanhildr. A like age is probably to be attributed to resemblances in language, such as those shown by the hortatory addresses and the accounts of dragon-fights cited in the last chapter (p. 60 f.). The fact that these resemblances sometimes occur in stories relating to entirely different characters need not prevent us from believing that they spring ultimately from a common origin. It cannot of course be proved that the materials from which the heroic poems are derived were themselves always in poetic (metrical) form. In principle we must admit the possibility that they were transmitted from one generation to another in a more or less stereotyped form of prose narrative, such as we find later in the sagas of Iceland and Ireland. But in point of fact we have no evidence whatever for the cultivation of such traditional prose narratives among any of the early Teutonic peoples, whereas there is good reason, as we shall see shortly. 78 EARLY POETRY AND MINSTRELSY [CHAP. for believing that narrative poetry was both ancient and widely cultivated. In the first place we may note that English and German poetry down to the ninth century shared a common system of metre and that the FornyrSislag, which is used in most of the Edda poems, differed but little from this type, except of course that it was always arranged in strophes. The application of this common metre to narrative purposes can scarcely be re- garded as a recent innovation, for English and German poems frequently exhibit verses and half-verses of very similar con- struction. Thus in the Hildebrandslied speeches are generally introduced with the formula: Hadubrant gimahalta, Hiltibraiites sunu, which is almost identical with a formula used in Beowulf: Wiglaf ina^elode, Weohstanes smiii. In the same German poem (v. 42) we find the verse : dat sagetun mi seolidante ('it has been told me by mariners '), with which we may compare Beow. 377 : \onne saegdon \aet saeli^ende. Note should also be taken of such phrases as (v. 55) ibu dir din ellen taoc ('if thy prowess is sufficient ') and definitely poetical expressions like (v. 43) inan wic fiirnam ('war carried him off'), as compared with Beow. 572 ^onne his ellen dealt and 1080 wig ealle fortiam Finnes \egnas. The number of such parallels might be greatly increased if we were to take into account passages from religious poems, especially the Old Saxon Heliand. For a very much earlier period direct evidence is furnished by the Roman historian Tacitus, who says {Germ. 2) that the Germani possessed ancient poems or songs (earmina) even in his time and adds expressly that they had no other means of preserving a historical records That these poems were partly of what may be called a 'heroic' character is clear from another passage {Ann. II 88), where it is stated that Arminius was still a subject of poetry among barbarian nations {ca7iitur...adhuc barbaras apnd gefites). In both cases the reference is in all probability to the peoples of western Germany rather than to the Goths. On the whole then we need not doubt that the heroic ^ celebrant carniinibiis antiquis, quod uniim apud illos memoriae et annalium genus est, etc. V] EARLY POETRY AND MINSTRELSY 79 poetry which we find in England and Germany during the seventh and eighth centuries had a long history behind it. Of course as to the form of the poetry current in the first century we are entirely without information. Many scholars hold that it was exclusively choric, not only in Tacitus' time but for four or five centuries later. This is one of the questions which we shall have to bear in mind in the course of our discussion. The earliest historical reference to the cultivation of poetry, or rather perhaps minstrelsy, in England occurs in Bede's account of the poet Caedmon {Hist. Eccl. IV 24). In this story we are told that it was the custom that, when the villagers met together to drink and amuse themselves, everyone should take his turn in singing to the harp. Caedmon, who had never been able to learn a song, used to leave the festivities and make his way home as soon as he saw the harp coming in his direction^ No information is given as to the character of these songs. Probably they would as a rule be quite short — perhaps not much longer than the hymn learned by Caedmon from the angel, which contains only nine verses. It is not unlikely that they resembled some of the metrical riddles more than any other form of Anglo-Saxon poetry which has come down to us. Longer songs, of a narrative type, may of course have been known. But it is a question whether such songs would deal with heroic themes or with folk-tales. We may think of the Scandinavian story of Svipdagr and MengloS, which is preserved in a number of different versions, ranging in date probably from the tenth century to the seventeenth or later. But we have already seen (p. 41 f.) that in the eighth century at least the recitation of heroic poetry was by no means un- known. Indeed we may infer from the language used by Alcuin {ridentium turuam m plateis) that it enjoyed a good deal of favour with the general public. Another of his letters^ speaks 1 nil carminum aliquando didicerat. unde nonnunquam in conuiuio cum esset laetitiae causa decretutn ut omnes per ordinem cantare deberent, ilk ubi adpropinquare sibi citharam cernebat surgebat a media caena et egressus ad suam domum repedebat. 2 Mon. Germ., Epist. Carol. Aeui, li 21 (ad Hygbaldum episc. Lindisfarnensem) : lectionis studium exercete. audiantur in domibus uestris legentes, non ludenUs in platea. 8o EARLY POETRY AND MINSTRELSY [CHAP. to much the same effect. We may gather from the use of the word citharista that these recitations also were accompanied by the harp. But the language of the letters seems rather to suggest that the performers in such cases were persons who made minstrelsy more or less of a profession. A century earlier the existence of professional minstrels may be inferred from the well-known story of Aldhelm' — that he used to take up his position on a bridge, like a professional minstrel {quasi artem cantitandi professum), and sing to the people in order to call them back to church. We are not informed as to the character of the poems he recited, but clearly they were of a type calculated to attract the country people. On the Continent we find very similar evidence. A passage in the Annals of Quedlinburg, to which we have already alluded and which dates perhaps ultimately from the tenth century, states that the country people used to sing of Dietrich von Bern I From a much earlier period we have an interesting reference to a Frisian minstrel named Bernlef, who became a disciple of St Liudger. He had been blind for three years, when he was brought to the missionary, but " he was greatly loved by his neighbours because of his geniality and his skill in reciting to the accompaniment of the harp stories of the deeds of the ancients and the wars of kings^" This incident appears to have taken place before 785. That minstrelsy was definitely recognised as a profession among the Frisians is shown by the last clause in the Lex Frisonum, which fixes a special compensation for injury to the hand of a harpists From all this we gather that in the eighth century there existed both in England and Germany a class of minstrels whose practice it was to play the harp and recite heroic poetry ^ William of Malmesbury, Gesta Pond/., v § 190 (from King Alfred's Handboc). * Aviulung Theoderic dtcitur...et iste fuit Thideric de Berne de quo cantabant rustici olim (Mon. Germ., Scr. Ill p. 31). ^ Et ecce illo disciimbente cum discipulis suis oblatus est cecus uocabulo Bernlef qui a uicinis suis ualde diligebatur eo quod esset affabilis et antiquorum actus regumque certamina bene nouerat psallendo promere, etc. Vita S. Liudgeri, II ( (Mon. Germ., Scr. II p. 412). ■* Qui harpatorem qui cum circulo harpare potest in manum percusserit componat illud quarta parte maiore compositione quam alteri eiusdem conditionis komini, etc. (Mon. Germ., Leg. in 699 f.). V] EARLY POETRY AND MINSTRELSY 8 1 in the village-streets or on bridges or wherever they could gather an audience. Now if we turn to the poems themselves we find that they also contain references to professional minstrels ; but these appear to have been quite a different class of persons from those with whom we have been dealing. At the close of his elegy (v. 35 ff.) Deor gives the following account of himself: "With regard to myself I will say that formerly I was the bard (scop) of the Heodeningas and dear to my lord. My name is Deor. For many years I have had a good office and a gracious master. But now Heorrenda, a skilful poet, has received the domain which the king had before given to me." There may be some difference of opinion as to the precise meaning of the word londryht^, but it is clear enough that the poet had been a court-minstrel and that he had been supplanted in the king's favour by a rival. In Beowulf also we find mention of a person who seems to hold a similar position. The word scop occurs three times in this poem (vv. 90, 496, 1066) — always perhaps with reference to the same man. In each case he is represented as singing or reciting, and twice mention is made of the harp. On the last occasion, when he recites the story of Finn and Hnaef at the banquet (cf. p. 2), he is called Hrd^gares scop, which seems to imply a sort of official position. The case of Widsith is somewhat different. The poet is a traveller who prides himself on the large number of nations he has visited. He states also that he served under various princes by whom he had been handsomely rewarded. The poem ends with some reflections on the life of wandering minstrels ; but these verses may be a later addition, like the introduction. At all events in v. 94 the poet speaks of his return home, when he presented to his lord, Eadgils prince of the Myrgingas, a valuable ' ring ' which had been given him by Eormenric. This present was a reward to Eadgils for his kindness in granting the poet the land formerly held by his father. It would seem then that the poet is represented as a man of good position. Whether we describe him as a wandering ^ Cf. my Studies on Anglo-Saxon Institutions, p. 369. 82 EARLY POETRY AND MINSTRELSY [CHAP. minstrel or not, his occupation is clearly to be regarded as court-minstrelsy and different therefore from that of Bernlef and the harpists mentioned by Alcuin. Apart from these personal notices there can be little doubt that the heroic poems which have come down to us were of courtly and not of popular origin. In the first place we may note their strongly aristocratic tone. This may be appreciated from the fact that all the women mentioned in the Anglo-Saxon poems are of royal birth, while the men are either princes or persons, apparently of noble or knightly rank, attached to the retinues of princes. On the rare occasions when persons of humbler rank are referred to, their names are not mentioned. In Beowulf no name is given even to the court-minstrel. Again the poems frequently refer to details of court etiquette, with which they seem to be well acquainted. In the later German poems this feature must of course be attributed to the conditions of the twelfth and thirteenth centuries, when the poems were composed. But it is quite as marked even in Beowulf We may note especially the long and detailed account of Beowulfs arrival at the Danish king's hall and the conversation which the chamberlain holds with the king on the one hand and the visitor on the other, before the latter is invited to enter. The chamberlain's exact position is remarked, when he approaches the king and it is added that " he knew the custom of knight- hood ^" Other members of the court also have their position or duties described (vv. 500 f., ii6if., ii65f., 1794 ff.) ; but most of all the poet loves to picture the movements of the king and queen (vv. 612 ff., 920 ff., 1 162 f.). Then again it should be observed that persons of royal rank are very seldom spoken of with disrespect. The rare exceptions to this rule probably all refer to persons of a remote past, Eormenric, Thrytho and Heremod, and in the last two cases the reprobation is qualified in a very marked way. Moreover the ground of censure is invariably violence, cruelty or treachery. Of immoral or unseemly conduct we have no mention. Indeed, except in the story of Weland — which stands by itself in many 1 Cu^ he dugU^ \>eaw (v. 359). V] EARLY POETRY AND MINSTRELSY 83 ways, as we shall see later — such subjectiv "jeem to be studiously- avoided. More than this the Anglo-Saxon heroic poems are entirely free from coarseness of language, such as we frequently find in Saxo's history, and indeed from references of any kind which could offend even the most fastidious taste. In general the same remarks are true also of the German and Scandinavian heroic poems, though not in the same degree. But the gnomic and theological poems of the Edda show a wholly different tone, which at its worst (e.g. in Lokasenna) verges on bestiality. Lastly, we must not overlook the fact that in dignity and polish of style the heroic poems far surpass any narrative works which the English language has to offer for many centuries later. It has been remarked that the composition of epic poetry requires a more or less professional training, and in the case of such poems as Beowulf this is doubtless true, not only on account of its length but also because a very large vocabulary is needed for the constant interchange of epithets which is one of its chief characteristics, while the allusions with which it abounds point to the possession of much historical or traditional lore^ At the time when it was produced the knowledge and leisure necessary for such composition is scarcely likely to have been found outside the entourages of kings. On the other hand we have seen that minstrelsy of some kind was cultivated even by peasants in Caedmon's time. We can hardly doubt that such was the case to a higher degree in court circles. In Beow. 867 ff we find a ' king's thegn ' com- posing an account of the hero's adventure immediately after its occurrence, and utilising apparently by way of illustration the story of Sigemund. This person may be the court minstrel; but the identity of the two is scarcely certain. In a later passage (v. 2105 ff.), referring to the banquet after the fight with Grendel, we hear of the king himself taking his turn with the harp : " There we had poetry and music. The old Scylding (Hrothgar) related stories of old time out of his great store of information. Now the martial^ hero would lay his hands on the joyous harp, the instrument that makes good cheer ; 1 Cf. Brandl, op. cit., p. 981 f. * It is generally thought that all these sentences refer to the king. 84 EARLY POETRY AND MINSTRELSY [CHAP. now he would recite a poem, true but sad ; now a story of marvel would be related in due course by the magnanimous king. Now again, bowed with age as he was, the old warrior would begin to lament that he had lost the martial vigour of youth. His heart surged within him as he called to mind the manifold experiences of a long life." It is held that the reference here is to lyrical effusions rather than to anything in the nature of epic narrative^; but I am inclined to doubt if we are justified in totally excluding the latter. I would rather favour the view that the cultivation of minstrelsy, including narrative as well as lyric poetry, was a general accomplishment in royal households, and that the office of court-minstrel was an honour given to that member of the court who had attained the greatest proficiency in his art. The statements of the poems as to the prevalence of court- minstrelsy during the Heroic Age are fully confirmed by the testimony of contemporary Roman writers. Perhaps the most important reference is a passage in Priscus' account of his visit to Attila in the year 448. After describing the banquet given by the king to his guests he proceeds as follows^ : " When evening came on torches were lighted and two barbarians stepped forth in front of Attila and recited poems which they had composed, recounting his victories and his valiant deeds in war. The banqueters fixed their eyes upon them, some being charmed with the poems, while others were roused in spirit, as the recollection of their wars came back to them. Others again burst into tears, because their bodies were en- feebled by age and their martial ardour had perforce to remain unsatisfied." It will be noticed that this account bears a curious resem- blance to the passage from Beowulf which we have just quoted. Nothing is said as to the language in which the poems were 1 Cf. Brandl, loc. cit. - K. Miiller, Fragmenta Historicorum Graecorum, IV p. 92. ew ly evoixivri^ Si eatripai 5^5es a.vf)dr)to various persons of being able to change into wolves or bears. As instances we may mention the case of Sigmundr and Sinfiotli given in Volsunga Saga, cap. 8, and the story of Biarki 1 Epistola Synodi Carisiacensis ad Hludowicum regem Germaniae directa (Mon. Germ., Legum Sect, ii, Capit. Reg. Franc, Tom. ii, p. 432 f.). "^ nos autem illos uidimus qui usque ad nostram aetatem Jurauerunt, qui huic rei interfuerunt et nobis uiiia uoce ueraciter sunt testati quae audierunt atque uiderunt. 2 Texts D, E, F. This entry seems to come from the Northumbrian Gesta, which were probably composed not very long afterwards. The last entry which we cun trace is for the year 806. VlJ SUPERNATURAL ELEMENTS 1 29 discussed above. For the latter we have already suggested an explanation. But though the motif may not have been a common one in heroic poetry — as compared with sagas relating to the Viking Age — there can be no doubt that the belief in shape-changing goes back to a remote antiquity. It is of frequent occurrence in poems and stories dealing with the gods, while similar ideas are widely prevalent among primitive peoples at the present day. Among other supernatural characteristics may be mentioned that of invulnerability, through the use of magic which rendered all weapons harmless — a feature found in Beowulf (in the case of Grendel) as well as in later works. Often too heroes are capable of superhuman powers of strength or endurance, as in Beow. 377 ff., 544 ff.^ though many of these cases may be set down to mere exaggeration. On the whole however such characteristics are scarcely as prominent as they are in the heroic stories of other nations. The love of the marvellous is more strikingly displayed in Procopius' account of Britain {Goth. IV 20) than in any of the poems which have come down to us. In the first place he says that the whole country beyond the great wall (i.e. the Roman Wall) was inhabited only by snakes and wild beasts, and that if any man ventured there he would die at once from the pesti- lential atmosphere. Then he goes on to say that Britain was the dwelling place of the spirits of the dead, and describes in detail how certain people who dwelt on the Frankish coast ferried the souls across. As to the truth of this story Procopius himself expresses scepticism ; yet he states that he had heard it from numerous witnesses. It is scarcely permissible therefore to suppose that he had been victimised by a humorist. More probably he is reporting stories actually current among the Teutonic soldiery in the Roman army, which doubtless contained adventurers from many distant lands. In short we have here to ^ A curious light on the enormous strength ascribed to Beowulf is thrown by a passage in the Liber Monstrorum (cf. p. 24). It is there stated that Hygelac (Beowulf's uncle) was of such immense size that no horse could carry him after he reached the age of twelve. His bones were shown as a marvel to visitors. 130 SUPERNATURAL ELEMENTS [CHAP. VI do with folk-tales^ which had been localised in Britain and were believed to represent its condition truly at the very time when Procopius was writing. It is exceedingly remarkable that such stories should obtain credence at a time when, as we know from more than one source, there was quite a considerable amount of communication between Britain and the Continent. Indeed Procopius himself says that large numbers of English emigrants had recently settled within the Prankish dominions. In the course of this chapter we have seen that many of the heroic stories contain elements derived from myth and folk-tale. The distinction which we have drawn between the two categories is that only the former deals with definite — though unhistorical — personalities^ It is commonly held that myth is a necessary element in heroic stories ; but this is a question which we must reserve for discussion in the next chapter. Further, we have seen that the presence of supernatural elements does not necessarily mean that the stories in which they occur were composed or modified long after the events which they relate ; that, on the contrary, such elements are to be found in contemporary or almost contemporary narratives. They must be taken as faithful reflections of the beliefs and ideas of an uncritical age. But it is scarcely correct to regard these elements as the distinctive characteristics of heroic poetry. Their chief domain in reality is the folk-tale, a far more primitive form of composition, which without doubt was in existence during the same period. The truly distinctive characteristics of heroic poetry are rather those which differentiate it from the folk-tale. ^ The folk- tale represented by the second story may of course be derived ulti- mately from some ancient custom ; cf. Beow. 26 — 52. ^ Not, of course, personalities consciously invented by an individual brain ; these must be classed under fiction. On the other hand myth must be held to include per- sonifications of the heavenly bodies and natural phenomena — as (e.g.) in Gylfaginning, cap. 10 ff. (from VaO'ru'Snismal, etc.), and certain Lithuanian folk-songs ('Uainos') — in so far at least as such personifications are of popular origin. CHAPTER VII. MYTHICAL ELEMENTS IN THE HEROIC POEMS. We have now to consider the question whether myth is a necessary element in the formation of heroic poetry. It has been noticed that historical persons figure in many stories of the Heroic Age, while others do not contain a single character whose historical existence can be authenticated. These latter stories are believed to be wholly mythical in origin, though they may not show any supernatural features in their final form. But even in stories of the former type it is held that some of the characters are almost always of mythical origin, and that their association with historical characters is a secondary development — due to confusion or to poetic imagination. In the last chapter we put forward the view that Beowulf, the hero of the poem, has been confused with a mythical character of the same name, and that the adventure with the dragon originally belonged to the latter. It can scarcely be doubted that Scyld Scefing, the father of this earlier Beowulf, was also a mythical character. The only element in his story common to English and Scandinavian tradition is that he is regarded as the ancestor of the Scyldungas or Skioldungar, the Danish royal family, and all analogies suggest that he came into existence as their eponymus. The brief account of him given in the poem might, except in two particulars, be applied to almost any successful king of the Heroic Age. One exception relates to the story of the funeral ship, on which the dead king's body is sent out to sea. In spite of Prof. Olrik's doubts* ^ Danmarks Heltedigtning, p. 248 ff. With this subject I have already dealt in The Origin of the English Nation, p. 287 f. 132 MYTHICAL ELEMENTS IN THE HEROIC POEMS [CHAP. I cannot but think that this is a reminiscence of ancient custom. The other is the reference to his arrival as an infant, likewise by sea — a story told more fully of Sceaf in certain English chronicles. The only question here is between myth and folk- tale. The story may fairly be classed under the latter head, though I think its origin is ultimately to be sought in a ritual myth. Scandinavian authorities, apart from Saxo^ record nothing distinctive of Skioldr, except that he was a son of Othin and the husband of the goddess Gefion* — which again points to myth. In Scyld-Skibldr we have the case of an eponymous ancestor appearing in the introduction to a poem which deals largely with the fortunes of his descendants. But there is no evidence that his own deeds ever formed the subject of an independent heroic poem. It would be somewhat hasty therefore to use this case as an argument for the origin of characters who are brought before us in the main action of heroic poems. Next we may take the .story of Weland, as to the mythical origin of which nearly all scholars seem to be agreed. It has indeed a historical or semi-historical connection in the fact that Weland is represented as the father of Widia (Wudga), i.e. the early Gothic hero Vidigoia mentioned by Jordanes (cf. p. 27) ; but this is held to be a secondary element in the story. In its original form the story is believed to have dealt only with the incidents related in VolundarkvitJa, viz. (i) the adventure with the swan-maidens, (ii) Weland's imprisonment by Nithhad and his revenged Behind the story itself however there lies a wide- spread belief in the existence of a supernatural smith of this name. Several places in Germany (Westphalia and Holstein) are reputed to be the scene of his operations, while in this ^ Saxo (p. 1 1 f.) records several incidents of which we know nothing from other sources. He represents Skioldr {Scioldus) as a reformer of the laws, but not as the first king. ^ This is stated only in Ynglinga Saga (cap. 5) ; but the question to be asked is whether it is likely that such a combination would be invented in late times. ^ From Deor's Elegy and the picture on the Franks casket in the British Museum it is clear that almost all the main features of the second part of the story were known in England. Reminiscences of the first part occur in the medieval German poem Herzog Friedrich von Schwaben. VIl] MYTHICAL ELEMENTS IN THE HEROIC POEMS 1 33 country he is connected with the cromlech called Wayland Smith, near Ashdown in Berkshire. In its ultimate origin this belief is traced to the myth of a fire-demon. Certainly it is to be noted that the name Weland is of a very exceptional type — apparently participial in form. One can hardly help suspecting that it once had a definite meaning, though this cannot now be determined with certainty^ Now there can be little doubt that the adventure with the swan-maidens is derived from a folk-tale. In this part of the story there is no indication of a fire-demon, or even of a smith, while analogies for the incident are fairly common both in Teutonic lands and much farther afield. We may confine our attention therefore to the second and better known part — that which deals with Weland's imprisonment and revenge. It is manifest that this story departs very decidedly from the ordinary standard of heroic poetry — firstly in the fact that the hero is here clearly represented as a smith, and secondly in the cruelty, treachery and vindictiveness ascribed to the chief characters. These are features which would be in place either in myth or folk-tale. But we may note further that there are analogies for part of this story, just as much as for the incident of the swan-maidens. As an example we may take Saxo's account of the robbing of Mimingus by Hotherus (p. 70 f). Mimingus is a satyrus, i.e. clearly either an elf or dwarf, who dwells in a cave in an almost inaccessible forest. Hotherus surprises and binds him and then takes away his sword and a magical ring. A connection between the two stories is shown even in the name, for Weland's most famous sword is called Mimmingl For a more remote parallel we may compare the story of Loki and Andvari. Indeed the spoiling of a dwarf is quite a common motif in Northern tales, while at the same time such beings are constantly credited with extraordinary skill in metallurgy. The distinctive feature in the story of Weland, apart from the revenge, is that sympathy is on the side of the smith. It is the end of the story — where Weland (Volundr) rises into the air and flies away — that is supposed to point most ^ It is usually connected with O. Norse vel, 'contrivance,' 'artifice.' ^ Waldhere, i 2 f. ; Thi'Sreks Saga, cap. 23 etc. 134 MYTHICAL ELEMENTS IN THE HEROIC POEMS [CHAP. clearly to a fire-myth. But this feature cannot be traced except in the Norse version ^ Moreover here we have also the ad- venture with the flying swan-maidens, in whose case there is no suspicion of such a myth. Setting aside this incident the story is perfectly explicable as a folk-tale founded on actual experi- ence. There can be no doubt that in the Heroic Age — and indeed in much earlier times — princes were especially anxious to obtain slaves, whether foreigners or not, who were skilled in metallurgy. And it is by no means incredible that such slaves were sometimes lamed in order to prevent any attempt at escape — although, quite apart from this explanation, smith's work may be regarded as a vocation natural to the lame man, just as minstrelsy to the blind. Further, it is likely enough that servile smiths, when cruelly treated, would take any opportunity that presented itself of avenging themselves on their masters. For the murder of Nithhad's sons we have a somewhat striking historical parallel in Eugippus' Life of St Severinus (cap. 8). Feletheus, king of the Rugii, who were settled on the Danube in the time of Odoacer, had a young son who one day was entrapped by some goldsmiths in the queen's service. They threatened to take his life ; but the saint intervened and rescued the boy on condition of the smiths obtaining their freedom^ What seems to me to be really the most remarkable feature in the story is that a person in this position should come to be made the subject of heroic or semi-heroic poetry ; for it is plain enough from many sources, especially Saxo's History, that smiths were generally regarded with deep aversion. In Deor's Elegy Weland is said to be a more distinguished man than Nithhad ; in VolundarkviSa he is called a chief of the elves, while the intro- duction makes him the son of a king of the Finns. Yet, except in the late ThiSreks Saga, his father's name is never given, and none of our authorities credit him with possessing a following of his own. ^ In Thi'Sreks Saga, cap. 30, Weland flies away in a garment which he has made from feathers collected for him by his brother Egill. It is thought by some that the engraver of the Franks casket had the same story in mind, since a figure catching birds is represented behind the form of Beaduhild. ^ It is commonly held that this account has been influenced somehow by the story of Weland. If so it is a valuable illustration of the process discussed in p. 1 19 ff. above. But the view seems to me somewhat far-fetched. VIl] MYTHICAL ELEMENTS IN THE HEROIC POEMS 1 35 Now is there any real necessity for the assumption that Weland's relationship to Widia is a secondary development ? It is found in two of the three national versions of the story \ and hence dates back in all probability at least to the sixth century. It is not found in the Volundarkvit5a ; but then Widia is alto- gether forgotten or unknown in Northern tradition. Moreover there is a distinct reference to offspring of Weland and Beaduhild at the end of the poem. Once grant that the relationship is old and the reason for the heroic treatment of the story becomes obvious. It is merely the reflection of the son's fame upon the father. As Widia is never said to be the son of anyone else the probability is that he was supposed to be illegitimate, and that a story was soon current as to his being the offspring of a union between a princess and a bondsmith. In such a case there would be a natural tendency to the accumulation of material from folk-tales about his parentage. If this view is correct the story must of course come originally from the Goths or some neighbouring people. I cannot see that the Westphalian traditions are any more con- clusive than the Berkshire cromlech as to its original home. If Weland was a character of folk-tale and his name had at one time a definite meaning, these local traditions may have been quite independent of the heroic story. The real difficulty seems to me to lie in determining the amount of material from folk-tale contained in the latter. We need not entertain any doubt as to the adventure with the swan-maidens. But what about Nithhad and Beaduhild .'* The latter name is not obviously framed to suit the character or circumstances of the unfortunate princess. Nithhad might be explained more easily in this way ; yet a Gothic prince of that name is recorded by Jordanes (cap. 22). At all events there is no conclusive evidence in either case that these characters did not originally belong to the story of Widia, Another story which is believed to be of wholly mythical origin is that of HeSinn and Hogni. In Kudrun it appears as ^ Waldhere is the only English poem which mentions Weland as the father of Widia; but I cannot admit that there is any ground for supposing this poem to have had a different origin from the rest (cf. p. 57 f.). The variant forms Widia — Wudga may be explained by English sound-laws, while Ni^had^ whatever its explanation, occurs also in Deor's Elegy. 136 MYTHICAL ELEMENTS IN THE HEROIC POEMS [CHAP. the introduction to a much longer story, from which point of view it bears a superficial resemblance to the story of Scyld and possibly to that of Weland. But since this feature is peculiar to the German poem and the second story seems to be entirely unknown from other sources, we can hardly do otherwise than treat the story of HeSinn and Hogni as an independent narrative. This must have been one of the most popular stories of the Heroic Age, since it can be traced in England, as well as in Germany and the North. None of the characters however can be traced in any historical work^, and the time to which it refers is quite uncertain. The Northern version of the story contains a supernatural element in the endless battle which forms its conclusion. It is generally held that this is the oldest element in the story and that Hild, whose name means * war,' was really a valkyrie. The whole story then is to be regarded as a myth of ' unceasing strife between conflicting powers^.' But we may naturally ask whether it is truly scientific, when dealing with a story known from three separate national traditions, to regard as the original element a feature found in only one of the three. It may be urged of course that the reconciliation, which in Kudrun^ takes the place of the tragic ending found in the Northern version, rendered it necessary to drop the mythical element ; and again that we have extremely little information ^ Saxo (p. 158 IT.) connects the story with the reign of Frotho III (Fro^i the Peaceful), and this may be an ancient feature, as Fruote von Tenemarke appears as one of Hetel's chief men in Kudrdn. ^ " Ein Bild des unaufhorlichen, allgemeinen, aber nie entschiedenen Kampfes entgegengesetzter Machte, des Aufgangs und des Niedergangs, des Entstehens und Vergehens, des Seins und Nichtseins" (Miillenhoff, ZfdA. xxx 229). Prof. Sijmons [Grundriss, III yri, where this interpretation is quoted with approval) regards the story as ' tiefsinnig.' I confess the interpretation is too deep for my comprehension. A totally different view is taken in Panzer's Hilde-Gudrun, where the origin of the story is traced to a folk-tale (p. 250 ff.). It seems to me that this theory is open to somewhat the same objection as the other, namely that it is founded too much upon features peculiar to one or other version. At the same time I doubt whether Wate's original connection with the story can be properly inferred from Wids. 21 f. The influence of folk-tales is clear enough in both versions of the story, but I think it is secondary. 2 From a passage in Lamprecht's Alexander (v. 1321), a work of the twelfth century, it appears that Hagen (Hogni) was killed in the earlier German version of the story. VII] MYTHICAL ELEMENTS IN THE HEROIC POEMS 1 3/ about the English form of the story. The latter remark is certainly true ; but the little that we do know practically precludes the possibility of a mythical interpretation. Strictly speaking the passage in Deor's Elegy (cf. p. 56) is not a reference to the story at all, but a matter of fact statement by the poet that he had been in the service of the Heodeningas. Hence, in view of the fact that this passage — together with Widsith, V. 21 — is probably by far the earliest reference to the story which we possess, I cannot regard the mythical interpretation as anything more than an extremely doubtful hypothesis^ Thus far we have been dealing with stories which are supposed to be of entirely mythical origin. Now we have to consider certain cases in which elements undoubtedly historical are believed to be blended with myth. As examples of this type we may take the stories of Waldhere and SigurSr. In both cases the historical elements are practically the same. In the former case it should be mentioned that the mythical theory is by no means universally accepted. Those scholars however who do adopt this interpretation base their view upon a supposed connection between the story of Waldhere and that of He'Sinn and Hogni. The chief points of resemblance between the two are as follows : (i) The heroine is called in the one case Hiltgund (Hildegyth), in the other Hildr (Hild). (ii) Both stories deal with abduction (so-called) and then with fighting, (iii) The man (Waldhere, HeSinn) who carries off the girl has in both cases to fight with a man called Hagen or Hogni. (iv) In both cases the combatants have previously been friends — though strictly this feature applies only to the Northern version of the HeSinn story. Now the first consideration carries no weight at all ; for half the feminine names which occur in Anglo-Saxon heroic poetry contain the element -hild- (e.g. Beaduhild, Hildeburg), while in the Continental and Scandinavian authorities also they are extremely common. Again, the last consideration obviously has little validity, except when taken in conjunction with the other two. But these (the second and ^ For the endless battle there are a number of parallels; cf. Panzer, op. cit., p. 327 ff. (also Pausanias, i 32. 3, with Mr Frazer's note). 138 MYTHICAL ELEMENTS IN THE HEROIC POEMS [CHAP. third) points of resemblance are, it seems to me, altogether misleading. To begin with it is hardly correct to apply the term ' abduction ' to an escape of hostages, such as the story of Waldhere relates ; at all events the conditions have nothing in common. Then the fight which follows is not, as in the case of HeSinn and Hogni, with an aggrieved father, or indeed with a pursuing force of any kind ; it is an unprovoked attack made by a third party in the hope of plunder. Lastly, the part played by the person called Hagen (Hogni) is quite different in the two cases. In one he is the injured father who is wholly responsible for the fight; in the other he is a vassal of Gunther (Guthhere), who is only drawn into it, with great reluctance, through the obligation of avenging his nephew (Patufrit), who has already been slain. It is true that a different version of the story appears in ThiSreks Saga af Bern^ Here Hogni (Hagen) is represented as pursuing the fugitives on behalf of Attila, while Guthhere does not appear. Now it has been widely assumed that this version is an independent and more original form of the story than that contained in Ekkehard's work, in spite of the fact that it does not make its appearance till nearly three centuries after the latter — and probably nearly six centuries after the composi- tion of the Anglo-Saxon poem. But the lapse of time in itself provides a perfectly adequate explanation of such divergencies, especially if we bear in mind the unfavourable conditions under which the heroic stories were preserved in Germany during the early Middle Ages. As the stories gradually became forgotten two tendencies are constantly observable : (i) to connect stories or incidents which originally were quite distinct, (ii) within the individual story to lose sight of all except the outstanding characters and incidents. Hence it is only in accordance with what we might expect that two different sets of opponents of the hero should be confused. For a parallel we may compare the late North German ballad on Eormenric's death (cf. p. 9), which describes how Theodric with eleven companions broke into the king's castle and slew him. It is generally agreed that ^ There is also a Polish version which has several peculiarities of its own but shows no special affinity with the form of the story found in ThitSreks Saga. VIl] MYTHICAL ELEMENTS IN THE HEROIC POEMS 1 39 this ballad is due to confusion of some kind with the story of HamSir and Sorli\ who were likewise enemies of Eormenric, though not connected in any way with Theodric. The saga itself really contains evidence which points to an earlier form of the story agreeing with that given by Ekkehard. For the hero is called Valtari af Vaskasteini-, and there can be no doubt that this expression is to be explained by the rocky defile mentioned by Ekkehard (v. 490 ff.) in the Vosges {saltus Vosagiis) — on the confines of Guthhere's dominions. Further it is to be noted that the story is introduced as an episode in the relations of Theodric and Eormenric, and that the hero is represented as a nephew of the last named. All scholars are agreed that this is due to late combination — and no doubt rightly. Yet Waldhere is associated with Theodric and Eormenric also in a number of German poems which are quite independent of the saga, and consequently it is by no means improbable that these combinations both preceded and helped to bring about the disappearance of Guthhere from the story. In its earlier form the affinities of the story with that of HeSinn and Hogni are, as we have seen, scarcely worth con- sideration. Yet apart from this supposed connection there is no case for believing it to have a mythical foundation — except the assumption that myth is a necessary ingredient in every heroic story. Whether it is to be regarded as history or fiction is of course quite a different question and one which we shall have to consider later. Of all the stories of the Heroic Age probably none has been more frequently referred to a mythical origin than that which deals with SigurSr (Siegfried). It is held by the great majority of scholars that the Nibelungenlied and the corre- sponding Edda poems — or rather the earlier poems or legends on which both were based — came into existence through the ^ The early North German version of this story, represented by the Annals of Quedlinburg (cf. p. 37, note), apparently made Eormenric perish in the fight. * It is held by many that this name was originally connected with IVascono lani, an early German name for Aquitaine (Gascony) and that the introduction of the Vosges ( Wasgunberg) was later and due to the confusion of two similar names. But, if there has been any such confusion at all, chronological considerations render it far more probable that the transference was in the reverse direction. 140 MYTHICAL ELEMENTS IN THE HEROIC POEMS [CHAP. amalgamation of an essentially mythical story with historical traditions of Attila and the fall of the Burgundian kingdom ^ The original elements of the former are believed to have been as follows : A young prince is brought up by a cunning smith in a forest, away from his father's home. On reaching manhood he gains an immense treasure by killing a dragon ; also he releases a maiden by overcoming difficulties and dangers, by fire or water, which were insurmountable to any other person. These two adventures are connected by many scholars. Later, the hero falls into the hands of foes, who slay him and take for themselves his wife and treasure. It is held that this story was originally a myth of light and darkness — applying however to the course of the year as well as to that of the day. SigurSr himself is a ' light-hero ' and Brynhildr a ' sun-maiden ' whom he releases at the dawn, while the treasure represents the blossoms of summer which the light- hero likewise wins by destroying the dragon of winter. Then, in the evening or autumn, he has to yield to the powers of darkness or winter. The original name of these powers was Niflungar or Nibelunge, a name connected with Old Norse 7iiH, 'mist,' Niflheimr, 'Hades.' Their chief representative is Hogni or Hagen, who, like SigurSr and Brynhildr, belongs to the mythical elements of the story. Many scholars also hold that the powers which destroyed the hero and appropriated his wife and treasure were originally identical with those from which he had won them at the beginning ; and this view seems to be more or less involved by the interpretation given above, since day and night, winter and summer are constantly alternating with one another. Now it will be obvious at once that the story as thus reconstructed differs greatly from both the forms in which it has come down to us. Indeed the only original feature preserved in both versions is the slaying of the dragon by SigurtJr. But it is only in the Norse version that the hero gains the treasure thereby ; in the Nibelungenlied this is obtained by a different encounter, with two princes named Schilbung and Nibelung, while in the Seyfridslied it really belonged to certain dwarfs, ' Among the exceptions mention may be made especially of an interesting paper by Prof. Mogk in Neue Jahrbilcher, I pp. 68 — 80. VII] MYTHICAL ELEMENTS IN THE HEROIC POEMS I4I the sons of Nybling^ Again, only the Norse version records that SigurSr released Brynhildr — from a perpetual sleep with which she had been punished ; the incident is not connected with the dragon adventure. The Seyfridslied does relate that the hero rescued a maiden from a dragon ; but here the maiden is Kriemhild, whom the dragon has carried off from her home. Further, it is only in the German version that Hagen kills the hero. In the Norse version the actual perpetrator of the deed is Gutthormr, but the instigator is Brynhildr herself Lastly, neither version of the story makes Brynhildr or the treasure return to their former owners, although — in the German version only — the first owners of the treasure bear the same name as those into whose possession it comes after the hero's death. It appears then that the original form of this story has been greatly obscured in both versions. The explanation given is that, through confusion with a historical tradition, the Burgundian kings, Gunnarr (Gunther) and his brothers, have taken the place (as well as the name) of the Niflungar. Hence, in order to form a just estimate of the theory it is necessary to examine the various mythical characters separately. These are — in addition to the dragon — Brynhildr, SigurSr and the Niflungar, including Hogni. The evidence for believing that Brynhildr was originally a mythical character lies chiefly^ in the identification of her with the valkyrie Sigrdn'fa (cf p. 13), who is mentioned only in the Norse version. The identity of the two characters is clearly recognised in the Helreit5 Brynhildar and also in the prose authorities. On the other hand it is not recognised in Gn'pisspd, which is supposed to be a late work, while the other poems leave it uncertain. The evidence therefore on the whole is not very strong. ^ It is stated however that Seyfrid thought that it belonged to the dragon. Hence this story is often connected with the Norse version. ^ The only German evidence worth consideration is the fact that certain rocks in the Taunus and the Palatinate are called the ' bed ' or ' chair of Brynhildr ' {lectidus Brunnihilde, Brinholdestut) in medieval documents. But I do not see how these names can prove anything more than the popularity of the story. In all lands it is customary to adopt such names from remarkable characters, whatever their origin may be. We may think of the cave of Frederic Barbarossa at Berchtesgaden or the numerous places called after Robin Hood in England. 142 MYTHICAL ELEMENTS IN THE HEROIC POEMS [CHAP. In the case of SigurSr the evidence, apart from the valkyrie incident, rests upon his being the slayer of the dragon. But it is agreed that this part of the story must be connected with the similar adventure attributed by Beowulf to Sigemund (Sigmundr) ; so that the question at issue is whether the exploit was first related of the father or the son. We have already seen (p. 123) that the argument in favour of the latter based on the agreement of the Norse and German authorities is in reality misleading. Hence the balance of probability is in favour of believing that the incident has been transferred to SigurtJr from Sigemund. In the case of the Niflungar the evidence depends upon the interpretation of the name. The use of the name is certainly somewhat curious. In the Seyfridslied (Part II) the dwarfs are called sons of Nybling. In the first half of the Nibelungenlied Nibelung is the name of one of the brothers slain by Siegfried in his youth, while the people who become subject to him, together with the treasure, on the death of the brothers, are collectively called Nibelunge. In the latter half of the poem however the same name is applied to the Burgundians. In the Norse version Niflungar always means Gunnarr and his people (i.e. the Burgundians), except perhaps in the expressions arfi Niflunga, hodd Niflunga (in AtlakviSa), by which the prose authorities at all events understood the treasure which SigurSr had taken from Fdfnir. The explanation given for this double use of the name is as follows. Originally it belonged to the mythical enemies of SigurtJr, i.e. Hogni and his people — whether these were identical with the former owners of the treasure or not. Later, when Hogni became associated with the historical Burgundian kings (Gunnarr etc.), the use of the term was extended so as to embrace them also. But it is to be observed that the interpretation of the name Niflungar as ' children of mist ' or ' darkness ' is not free from difficulty. In the Edda it is twice written Hniflung-^, and on both occasions the H- alliterates, whereas alliteration with 71- is never found. This ^ Helgakv. Hund. i 48, Atlamal 88. In the former case the name is used quite generally, like Ylfingar in the same poem. In the latter Hnijlungr is the name of Hogni's son. VII] MYTHICAL ELEMENTS IN THE HEROIC POEMS I43 fact suggests that the original form of the name was Hniflungar and that the form without H- is due to later influence — pre- sumably on the part of scribes — from German sources, where of course the H- (before -n-) would regularly be lost at a much earlier date. If so the name cannot originally have had any connection with O. Norse nifl, etc. In all these three cases then the evidence for the mytho- logical interpretation of the story seems to be at best inconclusive. But we have yet to consider the case of Hogni ; and here it must be remarked that the demonic character of Hogni is quite essential to the mythological theory. In the Norse version Gutthormr is a mere instrument and the person really respon- sible for the murder is Brynhildr herself — a feature obviously incompatible with the interpretation which we are discussing. Hogni's mythical origin is as necessary for this interpretation as that of SigurSr or Brynhildr. In order to maintain the theory the mythical character must be vindicated in all three cases alike. Now it has been remarked that Hagen {Haguno, Hagano etc.) is not uncommon as a personal name even in quite early times. This is a curious fact if the name had such associations \ But there is a much more serious difficulty. It is altogether contrary to reason or probability to separate Hagen the vassal of Gunther in the Nibelungenlied from Hagen the vassal of the same Gunther in the story of Waldhere. His character in the two cases is certainly quite different. He is brave in both ; but in the former he is both faithless and cruel, whereas in the latter he is an honourable man who is reluctantly drawn by circumstances into a course of action of which he heartily disapproves. But this is precisely the character borne by Hogni in the Norse version of the story of SigurSr — a fact which is the more remarkable since this type of character is extremely rare in heroic poetry. The agreement between the story of Waldhere and the Norse version seem to me to render it overwhelmingly probable that the character which they ascribe to Hogni was that which he originally bore. ^ There was of course another heroic character of the same name ; but this does not meet the objection. 144 MYTHICAL ELEMENTS IN THE HEROIC POEMS [ClIAP. I am not arguing now to prove that Hbgni was a historical person. He may be of fictitious origin or even mythical (though the latter seems to me extremely improbable). But if so clearly he must have been taken either from the story of SigurSr into that of Waldhere, or from the story of Waldhere into that of SigurSr. Which of the two he belonged to originally is a question of minor importance. The essential point is that an earlier German form of the story of SigurSr must be the link between the Norse version and the story of Waldhere ; for there is no evidence that the story of Waldhere itself was ever known in the North. The conclusion to which we are naturally brought is that in this earlier German form of the story Hogni bore the same character which is attributed to him in the Norse version. This character however is of course totally incompatible with a demonic origin ; and here, it seems to me, the interpretation which we are discussing hopelessly breaks down. I cannot help thinking that the investigation of the whole story has been greatly prejudiced by the application of wro\ig methods. There can be no doubt that a story of some kind — in which the adventures of SigurtSr were already combined with those of the historical Burgundian princes — was in existence long before the date of the earliest extant records, and that from this story, whether it was embodied in a single poem or consisted only of a mass of lays or legends, both the Norse and German versions are ultimately derived. It seems to me that, before trying to ascertain the origin of the various elements contained in the story, the object should be to determine the main features of this common foundation. The way to achieve this end is surely not by arbitrarily selecting one feature from the Norse version and another from the German, but by bringing together all the various features which the two have in common. To carry out such a process systematically would be quite beyond the scope of this book, but a brief outline of the scheme may not be out of place here. First then we will take the part of the story relating to the hero's early adventures, which is preserved mainly in a different set of authorities from the rest. The chief German authority is the late Seyfridslied which, as we have seen, is really made VII] MYTHICAL ELEMENTS IN THE HEROIC POEMS I45 up of two different ballads, inconsistent with one another. The Nibelungenlied contains only allusions to this part of the story, the action proper beginning shortly before the hero's arrival at Worms, The Norse version is given in the trilogy, Reginsm^l, Fdfnismdl and Sigrdn'fumal, as well as in Volsunga Saga, which is derived from these poems. The account given in ThiSreks Saga is mainly a combination of the German and Norse versions, though it has one or two features peculiar to itself. In this part of the story the common elements are very few in number, (i) SigurSr kills a dragon; (2) SigurSr gains a great treasure. In the Norse version the two adventures are combined, but in the Nibelungenlied the treasure belonged to Nibelung and his brother who had quarrelled and who are both killed by the hero. It may be noticed that Reginn and Fdfnir are also brothers who have quarrelled over a treasure, and they too are both killed by SigurSr; but Fdfnir has become a dragon — or perhaps a reptile. The Seyfridslied, Part I mentions only the killing of a dragon (serpent), while Part II unites the acquisition of the treasure with the killing of a dragon — a fiery dragon — but states that the hero erroneously thought that the treasure belonged to the dragon. Really it belonged to the three sons of the dwarf-king Nybling, who are friendly to the hero and not killed by him. As a further common element we may mention (3) that SigurSr is brought up by a smith. This story is found in the Seyfridslied, Part I, and in ThiSreks Saga — practically also in the Edda, since Reginn is represented as a smith. Again, (4) both in the "ballad and in Norse prose authorities SigurSr breaks the smith's anvil, though the circum- stances are quite different. It is doubtful whether we should connect the eating of Fafnir's heart, which enabled the hero to understand the birds, with the German story that he became invulnerable by bathing in the dragon's blood. Further, we have seen that the awakening of the valkyrie in Sigrdrifumdl has practically nothing in common with the rescue of the maiden (Kriemhild) from the dragon related in the ballad (Part II). Lastly, it is to be observed that though the hero's father has the same name (Sigemund) in all authorities, there is great discrepancy as to his childhood. In the Edda he is 146 MYTHICAL ELEMENTS IN THE HEROIC POEMS [CHAP. posthumous but knows his parentage, in the NibelungenHed he is brought up at his father's court, in the ballad. Part I, he leaves his home and goes to the smithy, in Part II and ThiSreks Saga he does not know his parentage — in the latter indeed he is a foundling and suckled by a hind. It will be seen that this part of the story is permeated throughout by the super- natural and marvellous. From the time of the hero's arrival at the Burgundian court we may take the Nibelungenlied for the German version, while the Norse one is best represented by the poems from the fragmentary SigurSarkviSa I to Atlamdl. For the earlier portion we have also to use Volsunga Saga and the prose Edda in place of certain poems which are lost (cf. p. 13), In this part of the story the elements common to the two versions are far more numerous and striking, (i) SigurSr comes to the Rhineland (Worms in the German version) and marries a sister (GuSrun, Kriemhild) of King Gunnarr (Gunther). (2) SigurtSr in super- natural disguise wins Brynhildr for Gunnarr. (3) SigurSr again in supernatural disguise sleeps with Brynhildr and takes from her a ring^ (4) Brynhildr quarrels with SigurSr's wife, and the latter shows her the ring^ (5) Brynhildr bitterly resents the treatment she has received and devises the hero's death. (6) SigurSr is killed by treachery ; but the versions differ in regard to the perpetrator of the deed. (7) The hero's widow is for a long time irreconcilable, but eventually is married to Atli (Etzel). (8) Gunnarr with Hogni and many others are invited to Atli's home, {gj The gold is sunk in the Rhine. (10) Gunnarr and Hogni are captured alive and the rest killed in Atli's land. (11) A demand for the gold is made and refused. (,12) Gunnarr and Hogni are killed. It will be seen that the supernatural is here confined to (2) and (3); indeed these are almost the only incidents in which it occurs in this part of the story. There is a difference between the two versions in regard to the character of the supernatural disguise. In the Norse version, where the two incidents are combined, SigurSr and Gunnarr have exchanged forms ; in the German Gunther is present in both cases, though Siegfried, who ^ On both these occasions the Nibelungenlied mentions also a girdle. VII] MYTHICAL ELEMENTS IN THE HEROIC POEMS I47 has rendered himself invisible, is the real actor. Several other important differences between the two versions have already been noted (p. 13 f.). In addition to these each version has of course many characters and incidents peculiar to itself. Of the discrepancies enumerated on p. 13 f. the fifth is by far the most important, since Kriemhild's revenge for Siegfried forms the central motif of the second half of the Nibelungenlied. In the Norse version no such central motif is to be found. In the prose piece Drdp Niflunga Atli's conduct is attributed to revenge for the death of Brynhildr, while in Volsunga Saga it is ascribed to his lust for SigurSr's gold. But in the poems them- selves no real explanation is given, and the connection between this part of the story and that relating to SigurSr is scarcely more than a personal one — viz. that GuSrvin, Gunnarr and Hogni figure in both. This however is a phenomenon for which parallels are to be found in other heroic stories, e.g. those of Beowulf and Weland. It is now held — and doubtless rightly — by the majority of scholars that the unity of interest imparted to the Nibelungen- lied by the motif of Kriemhild's revenge is a later improvement on the somewhat disconnected story given in the Edda. For our present purpose however the question is immaterial, since it is not contended that this part of the story is of mythical origin. In spite of the discrepancies noted above it cannot be denied that the two versions contain a remarkable number of identical features in this part of the story — a fact which renders all the more striking the very slight amount of agreement in the part dealing with the hero's early adventures. Unless all analogies are misleading the conclusion to which we are driven is that the original story began more or less where the Nibelungenlied begins, and that the hero's youthful adventures are later accre- tions, such as we see gathering round the childhood or ancestry of other heroes, e.g. Biarki (cf p. 120). We may add also the cases of StarkaSr, Hagen in Kudrun and perhaps Witege (cf. p. 135). They appear to be derived, in part at least, from folk-tales. One of these — affecting probably only the Norse version — may be identified with the Scandinavian story of Svipdagr and MengloS (cf p. 12), a variety perhaps of that of the Sleeping Beauty. Another is that of the forest dwarf 148 MYTHICAL ELEMENTS IN THE HEROIC POEMS [CHAP. who forges or preserves a magical sword. We may note that in ThitJreks Saga the smith is called Mimir, a name which recalls Saxo's Mimingus (cf. p. 133)'. The story of the treasure-guarding dragon may also be included in this category, though strictly perhaps it belongs rather to popular belief than to folk-tale. From the fact that some of these elements are common to both versions we may probably infer that the process of accretion had begun before the story reached the North. Yet there do seem to be some indications of a reflex influence^ — from the North or some region exposed to Northern influence — upon the development of the story in Germany. In addition to folk-tales we must take into consideration also a tendency which is often associated with them — the desire to account for an obscure name. This seems to be the most reasonable explanation of the names Nibelung and Nibelunge in the first part of the German epic and Nybling in the ballad — all denoting the original owners of the treasure. We have seen that in the Norse version, as well as in the latter half of the Nibelungenlied, Niflimgar means the Burgundians. May we not suppose that it was really a dynastic name^ like Scyldungas, Uffingas, Merewioingas} In that case of course hodd Nijlmiga {hort der Nibelunge) ought to mean the family treasure of the Burgundian kings. But is it quite certain that AtlakviSa does not use it in this sense .'' That it is identified with Fdfnir's treasure in later authorities may be due to subsequent German influence. As for the fact that the name Nibelmige is used for the Burgundians only in the second half of the German epic, may not this spring from some stylistic peculiarities of the 'common foundation'? It is not necessary to suppose that the latter was all the work of one author or even of one generation. In dealing with questions such as these we cannot hope to get beyond a reasonable hypothesis, since the paucity of common 1 It seems likely that Mimir was the dwarf's original name and that Saxo has given him a name which properly belonged to his sword ; cf. the phrase Hoddmimis holt (VafSr. 45), etc. * E.g. the name Schilbung and the references to Norway. The story of Sigemund and the dragon also belongs to a maritime region. 3 Cf. Skaldsk. 42 : " Gunnarr and Hogni are called Niflungar and Giukungar." VII] MYTHICAL ELEMENTS IN THE HEROIC POEMS 149 features between the two versions admits of few definite con- clusions. But from the time when the hero arrives at the Burgundian court the case is quite different. In spite of certain discrepancies there is no difficulty in determining the main out- line of the story. Even in the most important point of all — the true cause of the hero's death — the two versions are really in full agreement. Gutthormr does the deed in one version, Hagen in the other; but in both alike it arises out of the bitter resentment cherished by Brynhildr, owing to the deception which has been practised upon her. We have seen that this motif is incompat- ible with the current mythological interpretation of the story. But more than this, it is plainly not a motif derived from mythology at all, but from real life. It must not be overlooked that the Brynhildr and Hogni of the Norse version are in the nature of character-studies. Both appeal to our sympathies, though we do not approve of the actions which they commit or allow. Here we are in a region of thought as alien as possible to that of the folk-tale. But it is also alien to that period of thought, which was most open to the influence of folk-tales, the period which we have called Stage III in the history of German poetry. In such a period the person who destroyed the hero must necessarily be a villain as black as Hell. Between the instigator of the deed and the perpetrator, who by this time was Hagen — whether this was so originally or not is immaterial — the choice was made, not unnaturally in the circumstances, in favour of the latter, while the former was allowed, awkwardly enough, to drop out of the story. Thus the peculiarities of the German version may be explained quite naturally as modifications of an earlier form similar to the other — modifications necessitated by the conditions under which heroic poetry was preserved in Germany. The effect produced is somewhat similar to that which would be obtained by con- verting a modern problem play into a popular melodrama. The conclusion then to which we are brought is that the supposed traces of myth, so far as they have any foundation at all, are due to late accretions to the story, while the central motif in both versions alike is by no means of a mythical character, but essentially human. Consequently the story of I50 MYTHICAL ELEMENTS IN THE HEROIC POEMS [CHAP. VII SigurtJr stands quite on a line with the other stories of the Heroic Age. Most of them contain elements which may be interpreted as mythical ; but these elements are always most prominent in the latest forms of the story. It must not escape notice that those scholars who most strongly uphold the mythical interpretation base their arguments chiefly on such works as the Seyfridslied and ThiSreks Saga af Bern. The explanation is that myth is a growth which requires time to develop. Even Beowulf is no real exception to the general rule, for in the latter part the hero is probably confused with a namesake whose story may have been of considerable antiquity, while the only character in the poem who is quite clearly of mythical origin is the first ancestor of the Danish royal family^ ^ Cf. Schiitte, Oldsagn om Godtjod, pp. 35 — 38, where it is well pointed out that all the clearest cases of myth in early Teutonic records belong to stories dealing with the origin of nations or dynasties. "Den eneste udtrykkelige Myte, der udenfor specielt religLefse G^remal har vseret episk frugtbar i Folkevandringstiden er Ophavs- myten." CHAPTER VIII. THE USE OF FICTION IN THE HEROIC POEMS. The question how far the use of fiction was permitted in heroic poetry is of course one to which we cannot possibly hope to give a definite answer. All poetry which deserves the name claims to do something more than provide a bare record of facts. According to the ancient definition^ its proper function is to express the universal rather than the particular — what may happen or may have happened rather than what has happened. Some freedom of play for the imagination is therefore essential. These remarks hold good for early Teutonic poetry just as much as for Greek. If we could recover the poems recited in Attila's presence (cf. p. 84) we should doubtless find that they contained far more than a mere statement of facts. In the works which have come down to us however the degree to which freedom is allowed to the imagination varies very greatly from case to case. Thus in the poem on the battle of Brunanburh it is restricted within comparatively narrow limits, while in the almost con- temporaneous Hakonarmal the historical fact on which the poem is based is very largely obscured by a wholly fictitious narrative. We may naturally expect that the authors of heroic poems likewise differed in the treatment of their subjects, though not necessarily to the same degree. As an instance of a poem which obviously contains a large amount of fiction we can hardly do better than take the Anglo- Saxon Widsith. The greater part of this poem consists, as we ^ Aristotle, Poet, ix: 6 7ap iffropiKbs kuI 6 ■n-onjTr)S...Si.a iirTairv\(fi Qt^/Stj, Ka5/i7;f5t yalr), CjKfcre fw.pva/j.ei'ovs /j.rjXui' iveK Oidnr68ao, Toiii Si Kai fv vri^aaLv iiirep fiiya Xalrixa 6a\6.(i(nj% is Tpolrjv (l7a7d;i' 'EX^vt/s ^veK rjVKdfioio. k. t. X. In this passage the word ^pwj seems to have already begun to acquire its later meaning, viz. a distinguished man of the past (generally of the Heroic Age) who was honoured with worship, though not as a god. For such worship Teutonic records naturally furnish few parallels, since most of the Teutonic peoples became Christian either during the Heroic Age itself or soon after. We may compare however what Jordanes (cap. 13) says of the Goths: proceres suos, quorum quasi fortuna uincebant, non puros homines sed semideos, id est Ansis, uocauerunt. In Old Norse the name CEsir {*ansiz) is applied only to the gods (Othin, Thor, etc.) ; but we do hear occasionally of worship paid to heroes of the Heroic Age, as well as to distinguished persons of later times. An instance of the former (in the case of Hrolfr Kraki) occurs in Yngl. S. 4I. IX] THE HEROIC AGE OF GREECE 173 history abounds with references to the Heroic Age, and even Thucydides refers to it not unfrequently, though in a more critical spirit. In later times we have to notice especially antiquarian writers such as Strabo and above all Pausanias. The last-named derived his information very largely from local tradition and consequently the stories which he gives may often be independent of the poems. We may now consider briefly the chronological aspect of the Greek Heroic Age. It has already been mentioned that a passage in Hesiod's Works and Days speaks of an age of heroes inter- mediate between the bronze and iron ages, and that it further defines these heroes as those who fought at Thebes and Troy. To the latter number belong no doubt the various characters of the Iliad and Odyssey and the other poems (Cypria, etc.) which dealt with the Trojan cycle of legend, while the deeds of the former must have been treated in the Thebais and the Epigonoi. In the surviving Attic dramas which deal with the Heroic Age the distribution of subjects is as follows. Sixteen plays (three by Aeschylus, three by Sophocles and ten by Euripides, including the Cyclops and Rhesos) treat of the heroes of the Trojan war or their children ; six plays (one by Aeschylus, three by Sophocles and two by Euripides) deal with the Theban story ; and six plays (one by Sophocles and five by Euripides) are concerned with the doings of Heracles, Theseus or lason. It is to be observed that the heroes of the Theban story are always repre- sented as belonging to the generation immediately preceding that of the heroes of Troy, while Heracles, Theseus and lason are all loosely connected with one another and made roughly contemporary with the Theban heroes. The remaining three plays (Aeschylus' Suppliants and Euripides' Ion and Bacchai), if we are justified in regarding them as heroic at all, refer to persons much farther back in the genealogies. It appears then that the characters who figure most pro- minently in stories of the Heroic Age were, with few exceptions, ascribed to a period covering not more than three or four genera- tions. There are, it is true, a number of stories referring to much earlier generations — in addition to those treated in the 174 THE HEROIC AGE OF GREECE [CHAP. three plays mentioned above — but they seem to have been distinctly less popular than the others. On the other hand there is scarcely any reference to persons later than the children of the heroes who fought at Troy. With the evidence at our disposal it is impossible to fix any absolute dates for the Heroic Age. All that we can say is that the end of that age appears to coincide with the movement or series of movements, traditionally known as the Return of the Heracleidai, to which the Dorian states in the Peloponnesos were believed to owe their origin. According to the story, the Return took place in the second generation after the siege of Troy, and the grandsons of Agamemnon, the Achaean leader at the siege, were killed or expelled by the Dorians. Certainly it is to be noted that the scheme of tribal or political geography presented to us in the Homeric poems seems to show no trace either of Dorians in the Peloponnesos or of Ionic settlements in the eastern Aegean — another series of movements which are said to have been brought about by the Dorian conquest. The great majority of scholars apparently regard the story of the conquest as containing at least a nucleus of truth, though it refers to times long anterior to what we should call the historical period. The ancients themselves dated the events in question back to the twelfth or eleventh century (B.C.). But the evidence on which their conclusions were based is not of a very satisfactory character and will require careful consideration. Before entering upon this question it will be convenient to notice briefly the scenes of the stories and the localities and peoples to which the various characters belong. The scene of the Iliad is laid in the north-west corner of Asia Minor, a short distance south of the Dardanelles. But the stories introduced incidentally refer for the most part to places on the mainland of Greece, less frequently to localities in Asia Minor or Thrace. The distribution of the principal heroes is as follows: Agamem- non's territories, according to the Catalogue of Ships (II. II 569 fif.), lie in the north-east of the Peloponnesos, including the north- western part of what was later called Argolis and at least the eastern half of Achaia. Elsewhere (II. IX 149 ff., 291 ff.) he IX] THE HEROIC AGE OF GREECE I75 appears to have possessions in Messenia. His brother, Menelaos, rules over Sparta and other places in Laconia. Nestor's kingdom is on the western side of the Peloponnesos, to the south of Elis. Idomeneus belongs to Crete, Achilles to southern Thessaly (Phthiotis), Aias, the son of Telamon, to Salamis, his namesake to the eastern Locris, Diomedes to the eastern and southern parts of Argolis and Odysseus to the Ionian Isles, It must not be overlooked that most of these districts were of little or no political importance during the historical period and, further, that the territories of the kingdoms appear not to have coincided as a rule with the political divisions which we find in later times. The scene of the Odyssey is laid chiefly in the Ionian Isles, to a much smaller extent in the Peloponnesos. The wanderings of the hero himself appear to lie chiefly in regions to the west of Greece, though there may be reminiscences of the Black Sea. Some scholars relegate them largely or altogether to the realm of fairyland. Incidental references occur to Thesprotis (Epeiros) and the Aegean, as well as to more distant lands such as Egypt. Thebes was doubtless the scene of the lost Thebais and Epigonoi. The story of Pelops seems to have been connected chiefly with Elis and that of Perseus with Mycenae and Tiryns, while Minos belonged to Crete and Theseus to Athens. lason's home was in eastern Thessaly, but his story is largely taken up with journeys in the Black Sea and other distant regions. Heracles' adventures are spread over the greater part of Greece and many other lands, though Boeotia and Malis are perhaps the districts most prominent in his story. The scene of the Shield of Heracles is laid in Phthiotis. It appears then that the heroic stories are distributed over the greater part of the ancient Greek world. Certain districts however are excepted, and to these special attention should be given. In the first place we have practically no reference to Greek cities in Italy or Sicily or to heroes belonging to them, though we do hear occasionally of travellers' acquaintance with these countries. More important is the absence of any mention of Greek cities in Asia Minor^ and the adjacent islands, except ' In the Nostoi after the departure from Troy some of the Achaeans (Calchas, 176 THE HEROIC AGE OF GREECE [CHAP. those off the coast of Caria. The legends which speak of colonies led to Lesbos by Penthilos the son of Orestes or to Miletos and elsewhere by the sons of Codros are hardly to be reckoned among heroic traditions. The names indeed, at least in the first case, are taken from this source, but they form the subject of no connected story. Miletos is mentioned in the Trojan catalogue (II. n 868), but it is said to be in the possession of the Carians. To Chios there is only a geographical reference (Od. in 170 ff.), and though Lesbos is mentioned more frequently its inhabitants are treated as enemies by the Achaeans. The only real exceptions are the southern islands, several of which, such as Rhodes and Cos, send contingents to Agamemnon's army. Cyprus too seems to be fairly well known and its princes, though they take no part in the expedition, are on friendly terms with the Achaeans — a fact which renders the absence of reference to the Ionic cities all the more striking. In Greece itself nearly every district has a story connected with it. Attica however is one of the least prominent and possesses no hero of much note except Theseus. For the tribal distinctions which figure so prominently in later Greek history there is extremely little evidence in stories of the Heroic Age. The name AioXee? is not mentioned in the Homeric poems, while Awptee? occurs only once, as the name of one of the five peoples of Crete, and 'Idove^ once as that of a people (perhaps the Athenians) associated with the Locrians and Boeotians. On the other hand the most frequently used of all national designations is 'Ap^aiot, a name which in later times was borne only by the inhabitants of two comparatively unim- portant districts, Phthiotis and the north coast of the Pelopon- nesos. In the Homeric poems it appears to be a collective term for the inhabitants of Greece and the surrounding islands. In the same sense we find also Aavaoi, a name which later is used only in archaistic poetry. 'Apyeloi and "EXA.77j/e? seem properly to be geographical terms, though the former occurs frequently (the latter only once', in the form IlaveXkr]v€<;) as a synonym Leonteus and Polypoites) were made to arrive at Colophon. The story of Calchas' contest with Mopsos perhaps comes from the same source. 1 Once also, together with 'Axaioi (II. ii 684), as a name for the subjects of IX] THE HEROIC AGE OF GREECE 1/7 for 'Axaioi. Names of peoples, such as "A/Sayre?, 'ETretot, AiToyXoi, are of course frequently used ; but they denote com- paratively small sections of the nation. Though the term 'A^acok is used for the inhabitants of Greece collectively, it may of course really be the name of a tribe or people which was regarded as dominant at the time. At all events in Od. XIX 176 we find the Achaeans mentioned as merely one of five peoples which inhabit Crete. As an in- stance of a people who were apparently never included among the Achaeans we may take the Pelasgoi mentioned in the same passage. Here however we are faced with a question of nation- ality, for Herodotus speaks of the Pelasgoi of his own time as a barbarous people^ though at the same time he holds that several Greek peoples, especially the lonians and Athenians, were sprung from them. No indication is given in the Homeric poems that the Pelasgoi spoke a foreign language ; but this remark is true also of many Asiatic peoples, including the Trojans. On the question of Greek nationality there is unfortunately very little evidence either in the Homeric poems or in other stories relating to the Heroic Age. We cannot even tell whether the population of the Greek mainland was believed to be homogeneous. Only in the case of Crete is detailed ethnographical information given. In a passage cited above we Achilles, or rather Peleus. 'EXXds is used sometimes for a place or district in Peleus' kingdom, sometimes apparently in a wider sense. 1 He states (l 57) that in his time they inhabited Placia and Scylace, on the south coast of the Sea of Marmara, and KprjarQifa Tr6Xiv, probably in the Chalcidian peninsula (though some scholars emend this name to Rpdruva, i.e. Cortona in Tuscany). Down to the fifth century they are said to have also occupied Lemnos and Imbros (iv 145, v 26, vi 138 fif.), and in early times Samothrace (u 51), while their name was preserved at Antandros, in the Gulf of Adramyttion (vn 42). Later writers speak of the Pelasgoi as having formerly inhabited many other regions. Into the difficult problems connected with this name we need not enter here ; for the most recent and perhaps fullest discussion of the subject reference may be made to Prof Myres' paper in ihe/oitrn. Hell. St. 1907, p. i7off. As regards the etymology analogies indicate that \\.i\a.ayoi. represents an earlier form * Pelag-skoi. If this is a Greek word the most probable meaning is ' people of the sea ' (though another explanation has been proposed ; cf. Kretschmer, Glotta I 16 f.). But it may really be a national name (cf. ne\d7o;'es). In that case we may note that the use of the suffix •sko- in national names is Indo-European, though not Greek. 1/8 THE HEROIC AGE OF GREECE [CHAP. are told that this island contained five peoples, namely the Achaeans, Eteocretes, Cydones, Dorians and Pelasgoi. The first and fourth of these are well-known sections of the Greek nation, but we have no satisfactory evidence as to the nationality of the others. Herodotus (l 173) says that the Lycians came originally from Crete and adds that the whole of the island was once possessed by barbarians. Interesting light on the latter state- ment has been thrown by the recent discovery of certain inscrip- tions at Praisos, in what is said to have been the Eteocretan part of the island. These inscriptions are written in forms of the Greek alphabet which were current in the sixth and fourth centuries (B.C.) respectively, but the language is not Greek. Besides these numerous inscriptions dating from much earlier times have been found at Cnossos and elsewhere ; but they have not yet been deciphered. All that can be said at present is that we have no reason for discrediting Herodotus' statement. In Greece itself we have not such clear evidence for the prevalence of a non-Greek language. Here we are dependent on somewhat doubtful inferences from place-names. Yet the majority of scholars would not admit that the language was indigenous. Indeed the tendency at present is to believe that Greece and the Aegean islands were originally inhabited by peoples of one stock, the existence of which can be traced back in Crete for many thousands of years, and that these peoples were ultimately overwhelmed and absorbed — perhaps in the course of the second millennium — by invaders from the north. Asia Minor is supposed to have had a somewhat similar history. Originally it is believed to have been occupied by various kindred peoples, of which the most prominent were the Hittites of Cappadocia. Eventually — about 1200 B.C. according to the most recent view* — there took place a great irruption of Thraco- Phrygian peoples from the north-west, who became dominant throughout the larger part of the peninsula 2. 1 Meyer, S.-B. d. Akad. zu Berlin, 1908, p. 18. 2 Into the linguistic affinities of these various peoples we need not enter here. It will be sufficient to notice that the Thracian and Phrygian languages are commonly believed to have belonged to the eastern division of the Indo-European group. Certainly this is the case with the language of the Armenians, who according to Herodotus (vii 73) were an offshoot (dTroi/cot) of the Phrygians. The evidence of the IX] THE HEROIC AGE OF GREECE 1 79 In the Iliad the forces ranged in defence of Troy are drawn from a wide area, extending from the Axios (Vardar) on the west to Paphlagonia on the east and Lycia on the south. The nationaUties represented seem to be chiefly of Thraco-Phrygian stock, though a few, such as the Lycians and perhaps the Carians, belong to the indigenous population. On the other hand there is no reason for supposing that any of the peoples represented in Agamemnon's army were of other than Greek nationality. The story may therefore be regarded as one of national conflict. Yet it cannot be said that this feature is ever emphasised in the poems themselves, although the point of view throughout is that of an Achaean. For local or tribal patriotism the Homeric poems furnish us with little or no evidence. No Greek communities and few even of their princes are described otherwise than in terms of respect. We may point also to the old controversy regarding Homer's birthplace — a controversy which owes its very existence to the absence of any local patriotism in the poems. In this respect it will be seen that Greek heroic poetry agrees with Teutonic. We must now return to the consideration of the chronological problem. In ancient times, especially during the Alexandrian period, various attempts were made to calculate the exact date of the siege of Troy. Of these the most generally accepted was that of Eratosthenes, which was based on the length of the reigns ascribed to the kings of Sparta. This calculation brought the foundation of the Dorian kingdom at Sparta to the year 1104-3, a-"d eighty years were added to obtain the date of the fall of Troy. But it has long been pointed out that the figures given for the reigns of the early kings are so greatly above the average that they cannot be regarded with any confidence. Phrygian inscriptions themselves is unfortunately somewhat ambiguous. To the same eastern division belonged the ancient lUyrian languages, if the present dialects of Albania are descended from them. On the other hand the languages of the indigenous peoples throughout Asia Minor and the Aegean area are commonly believed to have been non-Indo-European. Yet Prof. Conway {British School at Athens, Ann. viii, p. 141 ff.) holds that the inscriptions of Praisos belonged to a language of this group. If this should turn out to be the case with the earlier Cretan inscriptions current views as to the early history of the Indo-European languages would require considerable modification. l8o THE HEROIC AGE OF GREECE [CHAP. Thus the first Olympiad (B.C. 776-5) was made to coincide with the tenth (or eleventh) year of Alcamenes and Theopompos. Alcamenes was the ninth in succession from Eurysthenes and Theopompos the eighth from Procles, and the number of years ascribed to the previous reigns amounts on the average to over thirty-five years for one dynasty and over thirty-nine for the other. But in kingdoms for which we have reliable information extending over a long period of time the usual average length is apparently between twenty and twenty-five years^. Hence, if the lists of kings themselves are to be trusted — and even this is very doubtful in view of the fact that they are almost identical with the genealogies — it is difificult to avoid suspecting that the total period ascribed to their reigns collectively is more than a century too long. More reliance is perhaps to be placed on the genealogies of the two royal families given by Herodotus (VH 204, vni 131). Leonidas (r. 488 — 480) and Leotychidas (r. 491 — 469), with whom we are on sure historical ground, are represented as fifteenth in descent from Eurysthenes and Procles respectively. According to all analogies therefore we should expect that the two latter flourished not very long before the middle of the tenth century. In other words the date given by Eratosthenes for the ' Return of the Heracleidai ' would seem to be from a century to a century and a half too early. A very reasonable explanation of the difficulty has been suggested by Prof Meyer who points out that several passages in Herodotus' history seem to imply the reckoning of a generation at forty yearsl Among the figures given we find (n 145) Heracles dated about 1330. Cleomenes, his descendant in the twentieth generation, was born about 530, or at all events not much later. For Eurysthenes and Procles, in the fifth generation from Heracles, this would give about 11 30, which is not very far from the date fixed by Eratosthenes. ^ For England from its unification under Alfred the Great to the present time the average is about twenty years; for France from 840 to 1793 it is between twenty- three and twenty-four years. ■^ Forsch. z. alten Geschichte, p. lyof. The reckoning is not due to Herodotus himself but taken over by him from an earlier writer. Prof. Meyer suggests as its author Hecataeus, who was a contemporary of Cleomenes. IX] THE HEROIC AGE OF GREECE l8l It is to be remembered in the first place that the date fixed for Eurysthenes and Procles is apparently that of their birth, and, secondly, that Eratosthenes' scheme is probably only a modification of a previously existing system, other varieties of which are quoted by Prof, Meyer^ Indeed it would not require any great exercise of ingenuity to point out traces of a more or less symmetrical distribution of the period covered by the reigns of the early Spartan kings'^ But, apart from any such specu- lations, we can hardly doubt, in the light of Prof. Meyer's showing, that the date for Eurysthenes and Procles is derived ultimately from a calculation based on the genealogies rather than from any contemporary written record or tradition. The genealogies themselves of course may represent tradition, so far as they are not interpolated*, but they point, as we have seen, to a much later date than that which we have been discussing^ If we substitute 32X 15 for 40 x 15, starting from the birth of Cleomenes, we are brought to about the year 1000. That must be regarded as the date really indicated by Spartan tradition for the birth of Eurysthenes and Procles. We may now turn for a moment to the genealogies of the other Heraclid families, namely those of Argos, Messenia and Corinth. The first of these places Pheidon in the sixth genera- tion, according to one version, in the ninth according to another, 1 Op. cit. p. 178 flf. ^ Eratosthenes reckons nearly 320 years from the accession (birth) of Eurysthenes to that of Alcamenes in the ninth generation, while the reigns of the father, grand- father, great-grandfather and great-great-grandfather of the latter make up 159 years. Sosibius, who reckoned by the Eurypontid dynasty, appears to have had a similar period of 320 years from Procles to Theopompos, although his dates were different from those of Eratosthenes — 1 091/0 to 771/0 according to Prof. Meyer {op. cit. p. 179 f.). The accession of Theopompos was equated with that of Alcamenes by Eratosthenes. Possibly these periods were originally sub-divisions of a longer period of 640 years, reckoned from the fortieth year of Cleomenes (or feeonidas?). ^ Two names (Prytanis and Eunomos) in the Eurypontid list are generally regarded with doubt, but none of the Agiad names is really of a suspicious character. The fact that Agis and Eurypon are not the first names in the genealogies ought not to be used as an argument against the trustworthiness of the tradition. In many Teutonic genealogies — e.g. the Gothic, Frankish (Merovingian), Kentish, East Anglian and Mercian — the name which performs patronymic function is not that which stands first in the list. * Prof. Meyer's view is not that the chronologists fixed too early a date for the Dorian invasion, but that the early parts of the genealogies themselves are unhistorical. 1 82 THE HEROIC AGE OF GREECE [CHAP. from Temenos, the uncle of Eurysthenes and Procles. Unfor- tunately different dates are assigned for Pheidon. The earliest, which is not generally accepted, places his reign about the middle of the eighth century. But even this, taking the longer form of the genealogy, does not carry us appreciably farther back than the Agiad list. The Corinthian genealogy places the last king, who is said to have been killed in 747, in the thirteenth generation from Heracles. This would agree with the longer form of the Argive genealogy ; several of the names however are generally regarded as suspicious. The Messenian genealogy is materially shorter. Apart from these Dorian genealogies there are some notices relating to the ancestry of persons belonging to other parts of Greece, which must not be ignored. Herodotus (n 143) states that Hecataeus, who was a prominent man at the beginning of the fifth century, claimed to be descended in the sixteenth (i.e. fifteenth) generation from a god. This probably takes us back to the Heroic Age, when divine parentage is common, whereas later it appears to be almost, if not entirely, unknown^ Again, it is believed that the genealogy of the Philaidai at Athens, which actually survives, though only in a corrupt form^ placed Philaios, the son of Aias, in the twelfth generation above Hippocleides, who was archon in 566. Further, according to Pausanias (l 11), Tharypas, king of the Molossoi, who was born soon after the middle of the fifth century, claimed to be descended in the fifteenth or sixteenth generation from Pyrrhos the son of Achilles. It will be seen that, though these genea- logies do not agree exactly, the discrepancy is not very great. They seem to indicate the existence of a belief that persons who flourished in the first half of the fifth century were removed by about fifteen generations from the Heroic Age. On the other hand Pindar {Pyth. IV 9 ff".) in an ode written in 466 and addressed to Arcesilaos IV, king of Cyrene, places that king's seventh ancestor, Battos I, in the sixteenth generation ^ Prof. Meyer {op. cit. p. 173 and note) cites the case of Telamon the son of Poseidon, ancestor of the priests of Poseidon at Halicamassos, whom he places after the Return of the Heracleidai. But the question is a complicated one. The genealogy cannot be used for our purpose, as we do not know where it ends. 2 Cf. Tdpffer, Attische Genealogie, p. 278 f. ; Meyer, op. cit. p. 174, note. IX] THE HEROIC AGE OF GREECE 1 83 from Euphemos the Argonaut, a contemporary of Heracles. This exceeds even the Agiad reckoning, for Pleistarchos, the representative of that family reigning in 466, was only in the twenty-first generation from Heracles. From the other non- Heraclid genealogies we should have expected that the number of generations to Arcesilaos would be about what is recorded for Battos l\ Whatever may be the explanation of this case, it will be seen that the other non-Heraclid genealogies are shorter than that of the Agidai by at least three generations — if we equate Philaios, Pyrrhos (Neoptolemos) and the grandson of Hecataeus' god with Aristomachos the grandfather of Eurysthenes. The dates which they indicate for the ' floruit ' of these persons are in no case earlier than the middle of the tenth century. As to the relative value of the two traditions we have nothing to guide us, and the same remark applies to the Greek genealogical evidence in general. Two points however must be insisted upon: (i) that the calculations of scholars of the Alexandrian age, or even earlier times, are not to be interpreted as evidence of tradition ; (ii) that the evidence of tradition, whatever be its value, brings the end of the Heroic Age at least towards the close of the eleventh century. Apart from the evidence discussed above, unsatisfactory as it doubtless is, chronological data for the Heroic Age itself seem to be entirely wanting. We know however that a highly advanced civilisation flourished in the Aegean in early times, and that it was succeeded by a long period in which both art and general culture were at a very low ebb. This latter period, which is commonly known as 'geometrical' from the type of art which prevailed in it, lasted, so far as one can judge, until about the end of the eighth century, at which time oriental influence began to make its appearance. The ' orientalising ' period again continued down to the beginning of the classical age. It is a common and natural hypothesis to equate the low-watermark of culture early in the geometrical period with the generations ^ Battos I is believed to have founded Cyrene about 630. It may be observed that the interval between that date and 466 is surprisingly short for the lapse of seven generations. 1 84 THE HEROIC AGE OF GREECE [CHAP. immediately following the Dorian conquest of the Peloponnesos. But unfortunately we cannot thereby obtain any certain date for the latter, since Greece appears to have had little contact with the outside world during the geometrical period. In recent years some advance has been made through the operations carried out by the British School at Sparta, which is perhaps the most important site for our purpose. From the stratification of the deposits Mr Dawkins, the director, has come to the conclusion that the earliest temple and altar at the sanctuary of Artemis Orthia date from the ninth or even the tenth century^ The temple, which must have been one of the earliest known, appears to have been a narrow and unpretentious structure of crude brick and timber. Some geometrical sherds were found beneath the floor, a fact which shows that the sanctuary had been in use somewhat earlier. If the sanctuary was founded at the beginning of the Dorian settlement at Sparta it is obvious that this result agrees well enough with the date indicated for the conquest by tradition. No relics of pre- geometrical times appear to have been found. When we turn back to the times of the earlier civilisation it is much easier to establish chronological equations ; for the presence of Egyptian objects among Aegean remains and of Aegean objects or representations of Aegean objects in Egypt shows that there was frequent communication between the two areas. Thus there is little doubt that certain Cretan remains date from periods contemporaneous with the twelfth and Hyksos dynasties. Others again clearly belong to the period of the eighteenth dynasty — at all events the earlier part of it. As to the date of the destruction of the Cretan palaces opinions still differ considerably; the most recent statement by Dr Evans is in favour of about 1350-. But this catastrophe did not bring the Aegean civilisation to an end. We find inscriptions dating from the subsequent period (Late Minoan III) apparently quite ^ British School at Athens, Ann. xiv, pp. 3, 18 f. ^ Cf. Hawes, Crete the Fore-runner of Greece, p. 18. A much later date is favoured by Dr Dorpfeld {Ath. Mittdlungen, xxxil 602), whose views on Cretan chronology differ greatly from those of English archaeologists. To this question we shall have to refer again in a later chapter. IX] THE HEROIC AGE OF GREECE 1 85 similar to those discovered in the earher stratum, and the various artistic types, though decadent, show no breach of con- tinuity. Indeed 'Mycenean' influence seems to have been more widespread (e.g. in Thessaly and Italy) at this time than in any earlier period. In Egypt vases of the same type are depicted in the wall-paintings on the tomb of Rameses III, who died about II 70. After this time however traces of Mycenean influ- ence are rarely found in that country. Within the last few years it has come to be noticed that the deposits dating from the last Mycenean period fall into two well-marked groups. The remarks made above, as to the art of Late Minoan III being a continuation of that of the preceding period, apply properly only to the first of tj[iese groups — repre- sented by the cemeteries of Zafer Papoura (Cnossos) and Phaistos, the late Mycenean megaron at Hagia Triada (also in Crete) and the late palace at Phylakopi in Melos. Other deposits, represented by the tombs found at Mouliana, Milatos, Kavousi and Erganos (all in Crete), though they have certain features in common with the former group, yet at the same time show a number of characteristics which are entirely new. Of these the most important are the practice of cremation and the use of fibulae and iron weapons, all of which are unknown in the Aegean before this time. From a careful study of the pottery found in these deposits Dr D. Mackenzie^ has come to the conclusion that it belongs to the same period as the famous ' Warrior Vase ' from Mycenae. This again is obviously con- temporary with a painted stele, likewise representing warriors, which was found in one of the latest graves in the lower town at the same place. Fibulae also were found here in the same group of graves. The importance of these observations lies in the fact that the armature of the warriors depicted on the vase and the stele corresponds in all essentials to what is described in the Homeric poems. This had already been pointed out by Prof Ridgeway^; but many scholars have attributed both objects to a much later period. Now however in view of the sequence ^ British School at Athens, Ann. XIII, p. 423 ff. 2 Early Age of Greece, p. 317. Representations of both the Vase and the Stele are given in the same work (p. 3135.). 1 86 THE HEROIC AGE OF GREECE [CHAP. which Dr Mackenzie has succeeded in tracing, in Cretan pottery and other articles, from the time of the destruction of the palaces onwards, it appears that the latter view can hardly be maintained. Lastly, Dr Mackenzie has pointed out that all the above deposits differ radically from those of the strict geo- metrical period found in cemeteries at Cnossos and Courtes and in the 'beehive' tomb near Kavousi (all in Crete). There are clear indications that all the latter belong to a subsequent time. As a result of his investigations then Dr Mackenzie has come to the conclusion that three well-marked periods can be distinguished in Cretan history, after the destruction of the palaces. But further, he believes that each of these periods coincides with a new settlement in the island — the true geo- metrical period with the Dorian settlement, the ' sub-Mycenean ' with that of the Achaeans, and the last true Mycenean period (Late Minoan III) with a settlement of Pelasgoi. The evidence of the deposits found at Sparta and elsewhere tends distinctly to favour the first of these identifications. For, though geo- metrical art was by no means confined to the Dorians, it may be presumed that their settlement was the latest of those which took place in Crete. Some scholars hold that they came there from the Peloponnesos, while others place their settlement in the island prior to the invasion of the peninsula — and ancient authority can be obtained for both views. But in either case it is improbable that the two events were separated by a long interval. Again, the identification of the second or sub-Mycenean period with that of Achaean settlement is rendered extremely probable by the resemblance which deposits of this period show to objects and customs described in the Homeric poems, e.g. in regard to armature and the use of fibulae and cremation. In the poems the Achaeans are clearly represented as dominant even in Crete, while in Greece itself, as we have seen, the exist- ence of other nationalities is practically ignored. In regard to the earliest of the three settlements Dr Mac- kenzie's theory may be open to more serious question. The ethnical affinities of the Pelasgoi are still quite obscure. Again, although new types, apparently derived from the mainland, do occur at this time, the break of continuity with the preceding IX] THE HEROIC AGE OF GREECE 1 87 age does not seem to be anything like so marked as in the subsequent periods. It is scarcely impossible that the destruction of the palaces may be due to naval warfare or piracy on a large scale, or even to commotions within the island itself For our purpose however this part of Dr Mackenzie's theory is of minor importance^ In other respects at all events the theory seems to provide a very satisfactory explanation of the phenomena. We have seen that the Spartan evidence, whether traditional or archaeological, affords no justification for dating the Dorian invasion very long before — or after — 1000 B.C. Now we find very good evidence for two distinct periods of culture between that event and the destruction of the Cretan palaces, which took place probably in the fourteenth century. The later of these periods is the one with which we are chiefly concerned ; for Greek tradition universally places the Heroic Age in times immediately preceding the Dorian invasion. It is therefore a fact of great significance that the deposits of this age agree in so striking a manner with the evidence of the Homeric poems. As to the relative duration of the two periods (Late Minoan III and sub- Mycenean or Achaean) archaeologists apparently have not as yet ventured to express an opinion. But it may be observed that the poems themselves give no indication that the Achaean dominion was believed to be of recent growth. The Cretan king Idomeneus is one of the oldest leaders at Troy, and his grandfather is said to have reigned at Cnossos before him. The evidence of the poems then favours the idea that even in Crete Achaean dominion lasted at least a century. This brings us back nearly to times when, fortunately, historical evidence is available once more, namely from the Egyptian monuments. During the thirteenth and twelfth cen- turies Egypt was threatened on several occasions by formidable armies. During the reign of Merenptah, probably about 1220, ^ Dr Mackenzie's theory would certainly gain in probability if it could be shown that the Pelasgoi were identical with the Pulesatha or Philistines (cf. p. 188). The arguments in favour of such an identification are obvious enough ; but they are scarcely of such a nature as to carry conviction. The appearance of -si- for -gsk- in the name is perhaps hardly to be regarded as an insuperable difficulty ; for we know nothing of the languages involved or of the sound-changes to which they were subject. l88 THE HEROIC AGE OF GREECE [CHAP. it was attacked by a host of Libyans and " foreign soldiers of the Libyans " whom " the miserable Libyan had led hither^" The names given to the confederates are Akaiuasha, Thuirsha (Turusha), Shakalesha (Shakarusha) and Shardina. Very early in the next century, during the reign of Rameses III, a fresh attack was made from the same quarter. A few years later Rameses encountered both by land and sea a great host coming from the north. "The Isles were restless, disturbed among themselves at one and the same time. No land stood before them, beginning from Kheta (Cappadocia and Cilicia), Kedi (the 'circling' of the Syrian coast at the Gulf of Iskanderun), Car- chemish, Arvad and Alashiya. They destroyed them, and assembled in their camp in the midst of Amar (Amurru; Pales- tine)^" The invaders here are called Shardina, Pulesatha (or Purusatha), Vashasha, Tchakaray (Zakar) and Danaau (or Danauna)^ Other Shardina appear to have been fighting on the side of the Egyptians. About a century earlier the Hittites brought a great confederacy against Rameses IL Among the names given here are Luka, Pidasa, Kalakisha, Dardenui and Masa^ On this occasion also we find Shardina in the Egyptian army. In the Tell-el-Amarna tablets, which date from shortly before the middle of the fourteenth century, we hear of Sirdana (apparently the same people) serving under the Egyptians in Palestine, and of attacks made upon the coast by Lukki, who are believed to be identical with the Luka'. Unfortunately scholars have not yet been able to come to any general agreement as to the identification of most of these names. It is commonly held that Pulesatha and Luka denote the Philistines and Lycians respectively. Many writers also identify the Shardina with the Sardinians and the Shakalesha with the Siceloi, but others connect these names with Sardis ^ Cf. Hall, The Oldest Civilization of Greece, p. i8of. ; Ann. of the Brit. School at Athens, vill, p. i8o; Petrie, History of Egypt, p. io8ff. 2 Hall, Ann. of the Brit. School, viii 183 ; cf. also Breasted, Ancient Records of Egypt, IV, p. 37 f. * Prof. Breasted (I.e.) gives these names as Peleset, Thekel (i.e. Zakar), Shekelesh (omitted above), Denyen and Weshesh. ^ Cf. Hall, op. cit. p. i77f. * Cf. Breasted, History of Egypt, pp. 336, 386 ; Hall, op. cit. p. i76f. IX] THE HEROIC AGE OF GREECE 1 89 and Sagalassos in Asia Minor. Other identifications which have received more or less assent are those of Akaiuasha, Thuirsha, Danaau, Dardenui and Masa with the Achaeans, Tyrrhenians, Danaoi, Dardanoi and Mysians, and of Vashasha and Pidasa with the inhabitants of Oaxos (in Crete) and Pedasos (in Caria) respectively. It has been supposed also that the Tchakaray, who are mentioned occasionally as mariners in later times, likewise belonged to Crete. In spite of the large element of doubt attaching to most of these identifications one important conclusion may be drawn with safety, namely that several of the nations mentioned had come from a considerable distance. Even those scholars who deny the references to Sicily and Sardinia hold that nearly all parts of the Aegean are represented in the lists. The inscriptions frequently speak of the invaders as coming from the sea or from islands. Thus the Pulesatha are said to be " in the midst of the sea." Again the king "slaughtered the Danauna in their isles ^" We find also the expressions " Vashasha of the sea," " Shardina of the sea," " Thuirsha of the seaV Such terms are said to be often used loosely; but under the eighteenth and nineteenth dynasties the territories of the Hittite kingdom in Syria and Cappadocia had become so well known to the Egyptians that it is incredible that any of the peoples of that region can be meant. The appearance too and the armature of the Shardina, as portrayed on the monuments, are quite incompatible with the supposition that they belonged to any of the countries round the south-east of the Mediterranean. We have seen that the northern invasion repelled by Rameses III was preceded by 'disturbances in the isles,' while the lands of the Hittites and their neighbours had apparently been overrun by the invaders before the attack upon Egypt. Now it has been noted that the great Hittite kingdom (in Cappadocia) appears to have been destroyed about the same time. In explanation of this the theory has recently been put forward' that the invasion repelled by Rameses III was closely ^ Cf. W. M. Miiller, Asien und Eiiropa, pp. 361, 363; Petrie, History of Egypt, III, p. 150. 2 Cf. W. M. Muller, op. cit. pp. 361, 371 ; Petrie, op. cit. pp. 151, 162. ' Cf. Meyer, S.-B. der Akad.zu Berlin, 1908, p. i8f. 190 THE HEROIC AGE OF GREECE [CHAP. connected with that irruption of Thraco-Phrygian peoples into Asia Minor to which we have already referred (p. 178). It will be seen that this theory has an important bearing on the Homeric question ; for in the Iliad we find the Thraco-Phrygian peoples already fully established in Asia Minor, and no hint is given that their settlement there was believed to be in any sense recent. But it is by no means impossible that the ' disturbances in the isles ' may refer to a displacement of population in a different quarter — which may or may not be connected with the Phrygian settlement in Asia Minor. The movement against Egypt was, in part at least, a maritime one, and when Rameses is said to have slaughtered the invaders in their islands the reference can hardly be to the old Hittite kingdom, which had long been known to the Egyptians. Surely it is more natural to connect the 'disturbances' with those national movements in the southern Aegean which eventually brought the Mycenean civilisation to an end. We have seen that the last period of this civilisation is believed to have begun in the fourteenth century and to have lasted some considerable time. The convulsions of Rameses' time (the early years of the twelfth century) may therefore mark a stage in the movements which brought about its destruc- tion. In any case it is from the Egyptian monuments of this period that we obtain the clearest evidence for contact between the true Mycenean civilisation and that ' sub-Mycenean ' or * Achaean ' type which followed it. The warriors of the invading forces^ are represented as armed with swords of the regular Mycenean pattern or with spears of no very great length. Some of them — the Pulesatha and Tchakaray — wear a peculiar head-dress, apparently made of feathers ^ which recalls the type used in later times by the Lycians according to Herodotus (vn 92). On the other hand the Shardina are depicted with very elaborate ^ A considerable number of the figures are reproduced in W. M. Miiller's Asien und Europa (cap. 27, 28) ; many also in the Histories of Egypt by Meyer and Petrie. 2 This head-dress is figured on a discus recently found at Phaislos (Crete) among deposits dating from 'Middle Minoan III' (cent. xvii?). The discus, if not actually of Cretan origin, is said to come clearly from some district under the influence of Cretan civilisation. Cf. Meyer, S.-B. d. Akad. zu Berlin, 1910, p. 1022 fif. I IX] THE HEROIC AGE OF GREECE I9I helmets^ which, except that they have no plumes, are almost identical with those borne by the figures on the Warrior Vase (cf. p. 185). The Shardina and many of the Pulesatha also carry round shields, held in one hand, as in the case of the warriors represented on the Vase^ and the Stele. This fact is especially noteworthy, since the round shield seems to have been totally foreign not only to the Egyptians themselves, but also to the Hittites and all neighbouring peoples^ while even in the Aegean area it was apparently not used in centres of Mycenean civilisation*. On the other hand both these features correspond to the type of armature described in the Homeric poems. It would not be correct of course to say that the portraits of the Shardina might be taken as faithful representa- tions of Homeric warriors. We find no trace of greaves, while the body-armour is of a less elaborate type than that described in the poems'. The Homeric type of armature — represented by the Warrior Vase as well as in the poems — belongs clearly to a later stage of development than the Shardina type, and therefore probably to a later age than the early part of the twelfth century. Yet there is sufficient resemblance between the two to render it more than likely that the one is descended from the other. Whatever may have been the causes which brought about the movement encountered by Rameses HI, some of the other references point distinctly to bands of mercenary soldiers, rather than to national migrations. This is especially clear in the case ^ They are said to be white ; but it is difficult to believe that they are not metal — possibly bronze overlaid with tin (cf. II. xxiil 560 ff.). 2 The shields figured on the Vase (apparently also those on the Stele) seem to have a section cut out of them ; but they can hardly be regarded otherwise than as round shields. Cf. Lippold, Munchener Arch. Studien, p. 406. ' Except the Assyrians (cf. p. 203, note) ; but there the evidence comes from much later times. * It occurs probably on the discus from Phaistos (cf. Meyer /. c.) and on a porcelain fragment from the third shaft -grave at Mycenae (cf. Reichel, iiher horn. Waffen, p. 58) — both times in conjunction with head-dresses of the Pulesatha or Shardina types — as well as on ivory objects from Enkomi in Cyprus (cf. Evans, Journ. Antkr. Inst., XXX 209), here also in conjunction with similar armature, though only the lower part of the helmet is visible. ' Many of the figures, both Shardina and Pulesatha, wear body-armour of some kind. Greaves are first found at Enkomi. 192 THE HEROIC AGE OF GREECE [CHAP. IX of the Shardina, who are mentioned for the first time as serving under the Egyptians in Palestine. But the earHest reference — in the Tell-el-Amarna letters — carries us back to the destruction of the Cretan palaces, if Dr Evans' date for this catastrophe is correct. It has been remarked that, in striking contrast with Mycenae and other early centres of civilisation in Greece, the Cretan palaces were almost entirely unfortified ; and the expla- nation commonly given of this fact is that their owners ruled the seas. If this is true we must conclude that the earliest maritime expeditions of the Shardina and their confederates did not take place without their consent. In the course of this discussion we have seen that, though historical evidence for the Greek Heroic Age is entirely wanting, later tradition points to the eleventh century as the time to which the poems and legends refer; and further, that, unsatis- factory as it doubtless is, this evidence is corroborated to a considerable extent by the results of archaeological investigation. In much earlier times various parts of the Greek world possessed a high civilisation, which has left remains of magnificent palaces and many elaborate works of art. The evidence of these remains does not correspond at all to the state of society revealed in the poems. But at the close of this earlier period many new objects — belonging, it would seem, to a new population — make their appearance ; and these latter do fulfil the conditions required. Lastly, it is of interest to note that from the fourteenth to the twelfth century Egypt and the Levant were frequently visited by bands of soldiers, who seem to have come from the Aegean or neighbouring regions and who outwardly bear a somewhat striking resemblance to the warriors described in the poems. It would appear that at this time the East must have been more familiar to the Greek world than Greek records would lead us to expect. This is the more noteworthy since during the following centuries the Greeks seem to have had but little contact with foreign nations. CHAPTER X. THE HOMERIC POEMS. It has been mentioned that, according to the theory now most commonly accepted, the Homeric poems were not the work of one author or even of one generation — that on the contrary they grew up gradually in the course of several hundred years, reaching their final form (in the case of the Odyssey) perhaps not before the middle of the seventh century. We must now review briefly the evidence on which this theory is based. The earliest historical references to the poems reach back only to the beginning of the sixth century, and even these are not altogether satisfactory. Cleisthenes, tyrant of Sicyon, according to Herodotus (v 6"]^ prohibited rhapsodists from reciting the Homeric poems, because they were full of the praises of Argos and its people. But the reference here is perhaps rather to the Thebais or Epigonoi than the Iliad. Again, it is said that when Athens and Megara were disputing about the possession of Salamis, both parties appealed to the authority of Homer in support of their contentions. If this story may be trusted it is of importance as showing that the poems were generally venerated at such an early date. And though all the evidence is late^, the form in which the passage in question (II. II 557 f) has survived does clearly suggest Athenian in- fluence. References in the works of other poets carry us back to a considerably earlier period. It has already been mentioned that Callinos is said to have attributed the Thebais to Homer, from which we may infer that poems under this name were already ^ Plutarch, Solon, cap. lo; Diogenes Laertius, Solon, cap. 48, etc. In some form or other the story was known to Aristotle [Rhet. i 15). 194 THE HOMERIC POEMS [CHAP. known. Archilochos, who likewise flourished before the middle of the seventh century, seems to have attributed the Margites to Homer^ The extant fragments of his works also contain several passages which apparently show Homeric influenced Terpandros, probably an older contemporary of Archilochos, is said to have invented a musical accompaniment for the Homeric poems I In Hesiodic poetry we find a number of references to heroic subjects. The Catalogue appears to have dealt with certain adventures of Odysseus, which were probably derived from the Odyssey*. This evidence, vague and somewhat uncertain as it is, renders it probable that Homeric poetry was in existence before the seventh century. Further than this we cannot hope for any direct external evidence, for the authors cited are the earliest of whom we know anything worth mention. Indeed the age of the various Hesiodic poems themselves is very problematical. It is customary now to attribute the Theogony and the Works and Days to the close of the eighth century and the other poems to the seventh. But Herodotus (ll 53) referred Hesiod, as well as Homer, to a period about 400 years before his own time, i.e. to about the middle of the ninth century. Turning now to the internal evidence, we may at all events regard one fact as established, namely that the subject-matter was determined at a period considerably anterior to those of which we have been speaking. It is not merely that the persons mentioned are uniformly referred to the Heroic Age, for we have yet to discuss the possibility that all the characters of this age are fictitious. But we have also to take into account the ethnographical indications contained in the poems. There is no hint of the presence of Dorians in the Peloponnesos, nor of the existence of Ionic states in Asia Minor or even in the Cyclades'; indeed the scheme of tribal or political geography which they present is far removed from anything we find even in the earliest ^ Archilochos, Fragm. 153 (in Bergk's Poelae Lyrici Graeci). ^ Cf. Croisset, Rev. des deux Mondes, 1907, 5, p. 605. 3 Plutarch, De Musica, iii 9. * Hesiod, Fragm. 65 f. (in Rzach's edition, 1902). ^ In Od. VI 162 ff. there is a reference to the sanctuary of Apollo in Delos. X] THE HOMERIC POEMS 195 records of the historical period. We shall see later that the same remark holds good with regard to the system of government depicted in the poems, e.g. in the universal prevalence of king- ship, and so also with their indications as to social organisation, religion and even ethical standards. Then there is the fact that in references to weapons bronze is far more frequently mentioned than iron; yet it is generally agreed that iron must have been in common use by the tenth century, while some scholars would refer its introduction to a much earlier date. Lastly, we may take note of certain passages and expressions which seem to contain reminiscences of the prehistoric civilisation of the Aegean — a civilisation which, as we have seen, had passed its zenith in the fourteenth century and which was probably altogether sub- merged in the convulsions which accompanied the Dorian and Ionic migrations. Among such reminiscences we may probably count the descriptions of the palaces of Menelaos and Alcinoos and the use of such terms as Kvavo<;. Of course it is not to be supposed that any part of the poems goes back to the period of the early civilisation. Yet the features noted above seem to me to point quite clearly to a time when some of the ancient palaces were still known and perhaps still inhabited. Hence it is not merely the case that the poems are concerned exclusively with characters of the Heroic Age. We are bound to conclude further that the environment in which these characters are placed is in general such as belonged to the same period. This of course involves the existence of a verbal tradition practically from the Heroic Age itself Indeed, we may say that it probably involves a poetic tradition, for we have no evidence for the existence of traditional prose narratives, whereas references to the cultivation of poetry in early times are fairly numerous. Now it is generally agreed that the Heroic Age — or perhaps we should say the type of civilisation and the ethnographical conditions with which this age is associated — cannot have lasted much beyond the close of the eleventh century. Hence the date accepted by many scholars for the completion of the Odyssey (cf p. 169) involves a period of more than three centuries, during which the Homeric poems were in process of formation. That is a long period for continued composition in one subject, and it 196 THE HOMERIC POEMS [CHAP. will be well now to review briefly the evidence on which this theory rests. Kirchhoff^ dated the 'later redaction' of the Odyssey between 01. 30 and 01. 50, or at all events not much before Ol. 30 (B.C. 656). To this conclusion he was led primarily by the reference to the voyage of the Argo in xn 59 — 72. It is probable enough that this passage implies acquaintance with a poem dealing with lason's adventures ; but the same can hardly be said of Kirchhoff's further suggestion, viz. that this lost poem must have been composed some considerable time after the colonisation of Cyzicos^ The accounts which we have of the travels and adventures of the Argonauts can scarcely be said to indicate that the earliest poems on this subject were composed at a time when the Black Sea was already familiar to the Greeks; on the other hand travellers from time to time may have penetrated into that region centuries before the foundation of Cyzicos*. Others have sought to show that the later parts of the poem betray an intimate acquaintance with the western seas, such as would be possible only after the development of Corinthian maritime enterprise about the close of the eighth century, whereas the knowledge of the same regions shown by the earlier parts is of the vaguest description*. Here the evidence is derived chiefly from the references to Sicily (St/ca/'t?;) and the ^ Die Composition der Odyssee, p. 85 f. ; Die horn. Odyssee^, p. 287 ff. ^ The references to the spring 'ApTaKl-q (cf. Od. x 107 f.) cannot be regarded as conclusive, since such connections are capable of more than one explanation — even if we bear in mind the name of the adjacent mountain ('Aprd/fTj). The mountain itself may have been known to the Greeks from early times. * In view of the evidence pointed out at the close of the last chapter one will do well to hesitate before denying the possibility of such distant expeditions in early times. But any communication which may have existed must have been interrupted by the invasions of the Bithynoi and Treres, probably in the ninth and eighth centuries. Note may also be taken here of what is said about the Cimmerioi in Od. XI 14 — 19 ; cf. Meyer, Geschichte des AlCerthiims, n, pp. 367 f. , 445 f. •* Cf. especially Wilamowitz-Mollendorff, Horn. Unt. p. 24 ff. The theory that the Ephyre of li 328 must be a different place from the Ephyre of i 259 seems to me very problematical if the author of the second book had only a vague knowledge of the geography of western Greece. Again, if Ilos Mermerides (i 259) is taken from the story of the Argo, is it really necessary that the source should be a different one from that referred to in xii 69 ff. ? X] TFIE HOMERIC POEMS 197 Siceloi, i^lybas (traditionally placed in the Gulf of Otranto, but perhaps rather a coined name) and Temesa, which is identified with Tempsa in Calabria. This theory seems to me to be open to much the same objections as the other. It is clear now that in prehistoric times the south of Italy was intimately connected with the eastern coast of the Adriatic, and we have no ground for denying that the former may have been known to Greek traders or pirates long before the date of the earliest colonies. A third argument, and one which has exercised a much wider influence, is based on the relationship of certain portions of the poems to the lost Cyclic poems (Cypria, etc.). It is held for instance that the Catalogues in II. Ii, at all events the Trojan catalogue, were taken from the Cypria, and again that in the Nekyia (Od. XI) and elsewhere use has been made of the Little Iliad and the Nostoi, as well as the Cypria. But, granting the correctness of these hypotheses, no conclusions as to date can be drawn from them unless the dates of the lost poems themselves are established. We have seen however that such is not really the case ; it is admitted that the attribution of these poems to Arctinos, Lesches and others does not occur until very late times. All that can be said is that they appear to have contained certain 'post-Homeric' features, such as purification for manslaughter (in the Aithiopis). With these we shall have to deal later. There is no need however to suppose that the lost poems were any more homogeneous than the Iliad and Odyssey. Again it is thought that certain passages betray the influence of Hesiodic poetry, while others indicate genealogies or relation- ships which are at variance with statements contained in the latter\ Among the former we may note especially the list of women in the Nekyia (Od. XI 235 — 327"), which is compared with the Hesiodic Catalogue of heroines. This evidence would be useful for chronological purposes if we knew (i) when Hesiod lived, (ii) that he was the first to compose catalogues of this kind. ^ From this it has been argued that the ' Odyssey ' known to Hesiod must have differed greatly from the poem which has come down to us. But it is to be remem- bered that there are quite as noticeable discrepancies between the Odyssey and the Iliad. ^ We shall have occasion later to notice more than one point in which this passage departs from the customary Homeric standards. 198 THE HOMERIC POEMS [CHAP. But unfortunately neither of these propositions can be admitted. To the first we have already referred. The majority of scholars hold that Hesiod cannot have lived much after the end of the eighth century. But there is nothing to show that he did not live before that time ; for no sound argument can be founded on the last verses of the Theogony. As for the Catalogue it belongs to a class of poetry of which the beginnings may go back to a remote antiquity. The presumption is that it originated in times when descent was still traced through the mother. On internal grounds many arguments have been brought forward for the purpose of showing that the poems in their present form have undergone a long process of development. With discrepancies in the narrative itself we need not concern ourselves. They are doubtless of importance for determining the question of single or composite authorship, but they do not necessarily point to authorship of quite different ages. For instance, one poet may have conceived of the Achaean camp as fortified, another as without fortifications. But that does not prove that the two poets were not contemporary, for it will not be disputed that the people of the Heroic Age were capable of building fortifications. Again, it may be that the original poem on the ' Wrath of Achilles ' did not originally contain Books n — VII. But, apart from one or two details which we shall discuss presently, there is nothing to suggest that these books are the product of an entirely different period. We may even take what is perhaps the most extreme case, that of the Doloneia. This book is joined on very loosely to what precedes, and its contents are practically disregarded in the rest of the Iliads Some critics even in ancient times seem to have believed that it did not originally belong to the poem. Moreover it contains a number of features peculiar to itself and several expressions which are regarded as indications of lateness. In particular there are some striking parallels to the passages in 1 Except probably in xiv 9ff., as has been ingeniously pointed out by Mr Lang {Homer and his Age, p. 276 ff.). In the same chapter Mr Lang shows that several features in the Doloneia which have been interpreted as marks of lateness may very well be due to the peculiar circumstances of the situation. 1 X] THE HOMERIC POEMS 1 99 the Odyssey, and it is held that in certain cases they are due to direct influence from the latter^ But, granting all this, we are still not in a position to decide whether the chronological differ- ence between the Doloneia and the earlier parts of the Iliad is to be reckoned at three centuries or two or one^. As no definite results can be attained from such considera- tions as these we will now confine our attention to arguments which are founded upon real or supposed differences of culture. The most important class of evidence for our purpose is that which relates to the use of the metals. Both bronze and iron are frequently mentioned and there can be no doubt that both were well known. But it has been observed that weapons are nearly always said to be of bronze ('x^oX.ko';, 'x^akKeov €y'x,o<;, etc.^), whereas iron is usually mentioned either as a substance or in reference to tools, especially hatchets\ Only in seven verses do we hear of iron weapons, even if we include in this category the knives mentioned in II. xvin 34 and xxni 30. Two verses (id. VII 141, 143 f) speak of an iron club and one (id. IV 123) of an arrow- head, while the other two (Od. xvi 294, XIX 13) refer to the arms in Odysseus' house collectively. The obvious inference from the statistics is that iron tools came into use before iron weapons, and though this was long thought incredible it has recently been shown that there is some ^ Gemoll, Hermes, xv 557 ff. (cf. xviii 308 ff.). Cf. also Shewan, Class. Quarterly, IV 73 ff., where this view is rejected. ^ Mr Lang {Homer and his Age, p. 265 ff.) has called attention to the fact that in V. 261 ff. Odysseus is represented as wearing a cap of a type which appears to have been in use during the Mycenean age. If it could be shown that the article in question was peculiar to that period, the lateness of the book would certainly be open to serious question. ^ Bronze is mentioned -279 times in the Iliad and 80 times in the Odyssey. In a large proportion of these cases the reference is to weapons. Cf. Helbig, Das homerische Epos, p. 329 ff. * Iron is mentioned altogether 48 times. In nine cases it is spoken of merely as a substance — a possession or article of trade. To these we may add fifteen more in which the word is used metaphorically as a standard of hardness, etc., and one (Od. IX 393) which refers to the testing of iron in water. Iron tools or implements are mentioned thirteen times, apart from the two references to knives given above. We hear also of iron chains (Od. i 204), the iron axle-tree of a (divine) chariot-wheel (II. V 723), and the iron door of Tartarus {ib. vili 15); cf. Cauer, Grundfragen der Homer-kritif?, p. 281 ff. 200 THE HOMERIC POEMS [CHAP. evidence for the prevalence of such conditions in Palestine \ Moreover a somewhat striking confirmation of the Homeric evidence was furnished by the excavations at Troy, where a small lump of unwrought iron was found among deposits belonging apparently to the fifth stratum-. It has been compared, and doubtless rightly, with the lump mentioned in II. XXHI 826 ff., though the latter must have been much larger. The presumption is that, like this, it was intended for some tool or agricultural implement ; for all the weapons found at Troy — few as they were unfortunately — up to the seventh stratum were of bronze. It is probable therefore that the use of the metals not only in the fifth stratum but also in the sixth — the great Mycenean fortress — was similar to that which is indicated in the Homeric poems. For those who believe that these poems are the work of a single author the words " Iron does of itself attract a man^" (Od. XVI 294, XIX 13) present a serious difficulty. I cannot believe in view of the evidence given above that iron weapons were regularly employed in the Heroic Age* and that the use of the word x^^x^^'^ is a piece of traditional poetic archaism. Other- wise however there is no alternative but to regard the Odyssey verse as an interpolation ^ But in reality it is no great step from iron knives and arrow-heads to the use of the same metal for spears or even swords. If we were to adopt the view that the age of the composition of the Homeric poems coincides with the period of transition between the first use of iron for cutting and piercing instruments to its general employment for weapons of all kinds, we should not necessarily require much more than a century for their development. Indeed the presence of iron ^ Cf. Macalister, Palestine Expl. Fund, Quart. Rep., 1903, p. 199; Lang, C/ass. Rev. xxil, p. 47. "^ Cf. Dorpfeld, Troja und Ilion, p. 368. ^ auros yhp i(pfKKfTai avSpa. alSrjpos. Cf. Burrows, The Discoveries in Crete, p. 2 1 4 ff. * Cf. Ridgeway, The Early Age of Greece, p. 294 f. Prof. Ridgeway allows the occasional use of bronze swords, e.g. in the case of Euryalos the Phaeacian (Od. viii 403 — 6). But the swords of Paris, Patroclos, Achilles and Odysseus (II. Ill 334 f., XVI 135 f., XIX 372 f., Od. X 261 f.) are described in very similar terms (^£0os dpyvp6rj\ov xaXfeoc). Further, the tendency of bronze swords to snap off short at the hilt is well illustrated by the case of Lycon in IL.xvi 338 f. ^ Cf. Lang, Homer and his Age, p. 192 f. \ X] THE HOMERIC POEMS 201 swords in graves of the sub-Mycenean period in Crete (cf. p. 185) would seem to show that the transition had begun within what may be regarded practically as the Heroic Age itself. Again, it is held that many anachronisms or chronological inconsistencies appear in the battle scenes of the Iliad. Some- times we find descriptions of armour and tactics which are thought to be copied from those of Ionic hoplites in early historical times and to be irreconcilable with the type of warfare depicted in other passages. Many of the chief men are repre- sented as armed with breastplates, for which, it is said, there is no evidence in the Mycenean age^ — though this statement is more than doubtfull In regard to the form of the shield also there is a discrepancy. Sometimes it is described as being of great length, 'like a tower' or 'reaching to the feet' — terms which would suit the long shields often depicted on objects of the Mycenean age. Sometimes on the other hand we hear of * round ' shields, which suggest the comparatively small circular shield of the historical period. These inconsistencies only concern us in so far as they are supposed to point to widely different ages. Unfortunately our information regarding the geometrical period is still very defective and we do not know how far its characteristics were the same in all parts of Greece. The same remarks apply with still greater force to the preceding or sub-Mycenean age. Hence, though it may be true that the round shield does not make its appearance in Attica till the close of the eighth century, we cannot argue from this that it was unknown in other parts of Greece^. In the more northern parts of Europe it appears to be the earliest form of shield which has been found. More important however is the fact that the Shardina and their confederates used the round shield as far back as the thirteenth century, while the shields represented on the Warrior Vase and the Stele from Mycenae (cf p. 185) can hardly be regarded otherwise than as a variety of the same type. Hence, whatever may be the connection between ^ Cf. Reichel, Ube7- homerische Waffen (Abh. d. arch.-epigr. Seminares d. Univ. Wien, Heft xi), p. 79 ff. 2 Cf. p. 191 and Y.\z.xis, Journal of the Anthropological Instiiuie, xxx 213 f. ^ Cf. Ridgeway, op. cit. pp. 324 f., 475. 202 THE HOMERIC POEMS [CHAP. the disappearance of the long shield and the growing use of body armour, there is not the slightest justification for sup- posing that the round variety was not used in the Heroic Age itself. The commonest type of Mycenean shield, the oval type con- tracted in the middle, belongs to a class of shields which occur in various parts of the world. It is probably akin to the Zulu shield,, though in this the lateral contraction has lost its meaning and almost disappeared. The primary purpose of the whole class appears to be for defence against missile weapons {€pKo66pa iroWa, dei^ai 8' rji'dyyetv (} irevdepi^ ocpp dir6\oi.TO. X] THE HOMERIC POEMS 21 3 yet been deciphered, and in view of the later inscriptions found at Praisos the probability as yet is distinctly against their being in the Greek language. But in any case they date from ages long anterior to Homeric times, and there is nothing to prove their continuity with the writing of the historical period. With regard to the passage in II. VI 168 ff. it is to be observed firstly that Proitos is one of the very earliest persons mentioned in the poems — some three generations removed from the characters of the Trojan story — and secondly that the curious phraseology seems rather to suggest that the poet was speaking of something which he did not clearly understand. On the whole then it is much to be doubted whether writing was a current and native practice during the period when the poems were composed. A reservation should perhaps be made with regard to the latest elements in the poems. Although definite evidence is wanting, the beginnings of the Greek alphabet may quite pro- bably go back to the ninth century, and it may very well be that the poems had not then attained their final form. But it should not be assumed that the alphabet was introduced simultaneously throughout the Greek world. Some districts may have acquired it generations before others, and Aeolis (including even Chios and Smyrna) was probably not one of the more advanced. Further, we must admit that in all probability its use was at first very limited. From all analogies we should expect that it was employed for inscriptions, correspondence, etc. for a very long time before it was made to serve any literary purpose. Such apparently was the case with the alphabet of ancient Rome and with the Runic alphabet almost throughout its history. Unless the conditions in Greece were quite exceptional we should not expect the alphabet to come into contact with heroic poetry for a considerable time. The theory that certain portions of the poems were written down from the beginning presupposes of course a use of writing quite different from that which is brought before us in the story of Bellerophon. This theory^ is, and necessarily must be, bound * The analysis of the Odyssey given by Prof. v. Wilamowitz-MollendorflF admittedly postulates a written text {ffom. Untersuch. p. 293). But I cannot assent to the 214 THE HOMERIC POEMS [CHAP. up with another theory, which we have already discussed, that these portions date from a period not earher than the seventh century. Our discussion has led us to the conclusion that the evidence for the latter theory is unsatisfactory. But, apart from this, it is inconceivable that the 'literary' portions, which are said to amount to several thousand verses, should contain no refer- ence, direct or indirect, to the use of writing. In Beowulf we find only one direct reference to writing (v. 1694 f) — an inscrip- tion in Runic letters such as had long been in use. But in the Christian additions or 'interpolations' we meet with three examples of the verb scrifan {forscrifajt, gescrifan — from Lat. scribere), which of course is indirect evidence for the use of writing, though in a different language^ If large portions of the Homeric poems were really of literary origin the authors could scarcely have failed to betray themselves by usages of this kind, even though they deliberately avoided all mention of writing. In addition to this general consideration we have to take account of the linguistic difficulties discussed above. We have seen that Hesiod's works show almost the same form of language as the Homeric poems, although the Boeotian dialect was quite different from anything spoken on the other side of the Aegean. Did Hesiod, who lived in the eighth century, probably before the Ionian conquest of Smyrna, really employ the ' impure Ionic' in which his poems have come down to us .-' He himself says (W. and D. 650 ff.) that he had never crossed the sea except (once apparently) to Chalcis. Presumably then his knowledge of heroic poetry was derived either from Boeotia, where he lived, or from Cyme in Aeolis, from whence his father had emigrated. But Cyme never came into Ionian hands. Are we to sup- pose then that either here or in Boeotia poets were already employing as their vehicle a form of language which according to the theory under discussion owed its existence (whether in proposition that the Catalogue of Ships in itself must come from a written source. This list scarcely differs in principle from the catalogues of Widsith. ■* Further indirect evidence, of native origin, is supplied by i\\Q vford facensta/as (O. Norse feiknstafir) in v. 1018, if the original meaning of this compound was 'harmful runes' (used magically). Cf. also vv. 317, 382, 458, 1753. X] THE HOMERIC POEMS 21$ ordinary speech or only in poetry) to certain political changes in a third district — changes too which had hardly begun much before the time of Hesiod ? This h\'pothesis seems to me quite incredible^ On the other hand if Hesiod's poems have undergone a change of dialect since their original composition may not the Homeric poems have passed through the same process ? In this case of course such a change would come about quite naturally if Smyrna or Chios was the original home of the poems. But here we must notice a curious feature in the 'epic dialect' to which we have already referred. Except in Aeolic forms — and this exception deserves to be remarked — the poems almost always preserve initial h-"-. This is a char- acteristic which the ' epic dialect ' shares with western or Euro- pean Ionic but not with the language of the Asiatic coastl Its presence raises a distinct difficulty in the way of supposing that the poems were lonicised in their original home. Now if the Homeric poems had been written down in Aeolic and preserved in literary form we can hardly doubt that they would have retained their original dialect, just like the works of Alcman, Alcaeus, and Sappho. It must not be argued that Ionic was the proper language of the epic, Doric and Aeolic of the lyric; for if the Homeric poems had become generally known in Aeolic nothing could have prevented this dialect from becoming the language of the epic also. We may assume then that they were not transmitted in written Aeolic. But the same argument really militates against the theory that they were written in Asiatic Ionic*, though in this case the difference of dialect is less striking. ' It may be added that we really know nothing of the Ionic of Hesiod's time. It is quite uncertain how far it had already developed those characteristics which we find in our texts. ''■ The few exceptional forms such as (r') oi5Xo»' (Od. XVII 343) may be due to fairly late scribes familiar with (eastern) Ionic texts. ^ We may note also the absence of the literary Ionic forms /cdre, kws, etc. * It is true that we do not know exactly when h- was lost. But before that change took place H cannot have been used for e ; consequently a wholesale fitT ay pa.tJi)jLa.T Iff yibi would be involved (doubtless also affecting the representation of ^), iust as in the case of Athens. 2l6 THE HOMERIC POEMS [CHAP. A form of language practically identical with that of the epics appears in the remains of several poets of the seventh and sixth centuries, such as Archilochos and Solon. The only noticeable difference is in the proportion of non-Ionic forms which they use. Now since these authors belonged to quite different districts we must, if we are to trust our evidence, infer the existence of a kind of literary language at this time. Its difference from Asiatic Ionic is of course comparatively slight, and in the texts which have come down to us it is not always carefully observed'. Still there is a sufficient amount of regularity to show that it was generally recognised. The true home of this literary language must not be sought in Athens, but rather in Euboea or the Cyclades; and we may probably attribute its spread, in part at least, to the influence of the poems of Archilochos. But it certainly affected the writing of Attic for some two centuries, and there can be little doubt that it was thoroughly domiciled in Athens at quite an early date — probably in the seventh century. I cannot see any objection therefore to supposing that it was in Athens that both the Homeric and Hesiodic poems acquired those peculiar lin- guistic characteristics which we comprehend under the term 'epic dialect.' This of course brings us back to the story that the poems were collected or written down by order of Peisis- tratos. The evidence for the story is late, and its truth is hotly contested by many scholars. But at all events it has the merit of providing a satisfactory explanation of the linguistic pheno- mena. Certainly, if we may judge from the analogy of other peoples, heroic poetry would not by any means be among the first species of literature to get committed to writing. There is no reason whatever for supposing that anything of this kind was written down in England before the eighth century, i.e. at least a century after the language was first applied to literary purposes, in Aethelberht's time. The same remark seems to be true of Ireland, Germany and the North, while Bosnian heroic poetry is being written down only in our own generation, and not by natives of the country. It is only natural therefore to expect ' Cf. Hoftmann, Griech. Dial, ni, p. 549 f. ; Fick, Neue Jahrbiicher, i 504 ff. X] THE HOMERIC POEMS 21/ that the Homeric poems would be written down according to an orthography which was already well established. This orthography no doubt represents more or less truly the form in which the poems were recited in Athens at the time. But does it also represent the form in which they were recited at (let us say) Sicyon ? That is a question upon which we have no direct evidence. But it is worth noting that in Doric and other non-Ionic states we find a number of ancient inscriptions in hexameters or elegiacs which contain epic words and forms. We may note especially the Gen, sg. ending -olo, e.g. KucriyveToio, oBoio, ApaOdoLo, and above all false imitations of epic forms, such as TXacTKiafro^. But in all these inscriptions Ionic char- acteristics are conspicuously absents Does this mean that wherever the heroic poems were introduced the rhapsodists tried to accommodate them, as far as possible, to the language of the district? But if so, is it possible that the Ionic element in our texts is wholly due to the rhapsodists } A Chalcidian or Naxian or Athenian audience would certainly experience at least as much difficulty in following a purely Aeolic Iliad as a southern English audience would have in listening to a purely Northumbrian or Mercian Beowulf. On the other hand an Iliad lonicised in Chalcis or Naxos would be easily intelligible in Athens^ ^ Cauer, Delectus Inscr. Graec. 83, 84, 91. We may compare also such epic expressions as evpvxopo, yaias awo irarpidos, noXvfjLeXo, /cXe/oj airdirov (id. 54, 83, 445, 202), and, more particularly, an inscription on a bronze discus from Cephallenia (cf. Cook, C/ass. /Review, xili 77 f.) : ExcoiSa fi aved€Ke Ai/oj (fopoiv fx.eya\oio XO-^Keop hot viKaae Ke. cit. 78) found near Caere. So also with the heroic names used by Pindar and other non- Ionic poets — not to mention the Latin forms. Yet these poets use the Ionic forms of foreign names, such as M^Soi ' (Cypr. Ma-to-i), which had come to them presumably through Ionic channels. On the other hand we find in inscriptions on Chalcidian vases more purely Ionic forms, e.g. Aivees {ib. 545), than those preserved in our text. These seem to count against any place except Athens as the home of the final form of the ' epic dialect.' * From the fact that Pindar and other non-Ionic authors use what is apparently an Ionic form — indeed, strictly speaking, a western Ionic form— in the poet's name 2l8 THE HOMERIC POEMS [CHAP. At all events we have seen that there is a very serious objection on chronological grounds to the view that Hesiod's poems were composed in a form of 'impure Ionic' borrowed from the Asiatic coast. In this case it is surely far more probable that the Ionic element is due to the rhapsodists, whether in Chalcis or Athens. But is there any real reason for denying that the Homeric poems may have had a similar experience? This is a question which I do not feel qualified to answer. But it seems to me to deserve more attention from scholars than it has received as yet. In this chapter we have seen that the Homeric poems con- tain elements of great antiquity. Although we have no means of fixing an exact date for these elements, we can hardly doubt that they originated at a time before iron had come into general use for weapons. According to the prevailing opinion of archae- ologists this innovation cannot have taken place after the tenth century. We need not suppose that any considerable portions of the poems in their present form date from such an early period. But the ' type ' must have been fixed by that time, and to a considerable extent also the subject-matter. Still more clearly we have seen that there is no ground for supposing that the earlier and later elements are separated from one another by a wide interval. For the idea that the earlier elements reflect the conditions of the last age of Mycenean splendour — probably about the thirteenth century — while the later elements betray acquaintance with conditions of the seventh century, we have not been able to find any justification. The period intervening between the Mycenean age and the beginning ("O/Miypos), while they give the names of the heroes themselves in non-Ionic form, we are justified in concluding that they had acquired the former from a different (presumably literary) source. Certainly the earliest references to the poet come from Ionic authors. Again, Thucydides (in 104) is clearly recording a generally accepted opinion when he quotes the Hymn to the Delian Apollo under Homer's name ; and I can see no reason for doubting the identity of Semonides' Xtos avf\p (cf. p. 209) with the author of this poem (v. 172: Tvl\ depdiruv KX^ea TrpoTip dvOpdiiruv i)/xvriv toIs TpaxrX (rvfji.fiaxr}(TdvT(t)v, and it is commonly believed^ that this is the list which has been incorporated in our text of the Iliad. Proclus' words, taken literally, would seem to indicate that the list did not include the Trojans themselves. The ' allies ' proper begin perhaps at V. 840, and consequently it has been held that vv. 816 — 839 are a later addition. Considering the nature of the evidence, this view can scarcely be regarded as beyond question. But there certainly are noticeable differences between the first and second portions of the list, apart from the fact that the former contains no national names except TpSxs and Aapbdvioi. The first part is much more dependent on the Iliad than is the second. Thus all the personal names which occur in it are to be found in other parts of the poem, while several passages (eight verses^ out of the twenty-four) have been borrowed practically verbatim. On the other hand, out of the twenty- five persons mentioned in the second part of the list eleven are not met with elsewhere in the Iliad, while four of the others are not elsewhere associated with the nations (Mysians, Phrygians and Meiones) to which they belong here. Again, though several verses show acquaintance with other passages in the Iliad, there is apparently no single verse which has been borrowed entire. The most striking fact however is that in two cases we have references to heroes who are said to have been slain by Achilles in the river. Neither of these persons is mentioned in the account of the river fight given in the Iliad. In explanation of this it has been suggested that another scene of this kind occurred in the Aithiopis ; but we have no evidence to this effect, and it is surely at least as probable that they are derived from a different version of the river fight described in the Iliad. ^ Cf. Kinkel, Epicorutn Graec. Fragmenta, p. 20. ^ See the Addenda. 3 V. 822 f. from XII 99 f., vv. 831 — 4 from XI 329 — 332, v. 838 f. from XII 96 f. The borrowing of vv. 831 — 4 can hardly be ascribed to the same man who took Amphios (the son of Selagos) from v 612. Ignorance of the contents of the poem such as we find here is very rare ; its occurrence therefore in what is probably a very late addition to the text deserves to be noticed. XI] THE TROJAN CATALOGUE 245 In both parts of the list we meet with a considerable number of geographical names — of cities, rivers and mountains — many of which (five in the first part and twelve in the second) do not occur elsewhere in the ^ Iliad. The majority of the names mentioned belong to the immediate 1 neighbourhood of the coast, and it is a fact worth noting that these i almost all fall into three distinct groups — in Paphlagonia, the Troad and Caria respectively. There is a great gap covering apparently the whole of the Bithynian coast and another embracing the coast of the Aegean from Troy itself to Mycale. In explanation of this fact it has been sug- gested^ that the names in the coast-districts are derived from an early j;, poem or poems on the voyage of the Argo. This hypothesis might certainly I account for the mention of the Paphlagonian names in v. 852 ff. But they are the only names in the second part of the list for which it gives any explanation ; for the suggestion that the references to the Carian localities in V. 868 f are connected in any way with the story of the Argo can hardly be taken seriously. Hence, if we are to trxist the hypothesis in any form, it is more probable that v. 853 ff. are a subsequent addition to the list. But even in the first part the evidence in its favour is of the most slender description 2. Of the thirteen place-names which occur in this section eight are found elsewhere in the Iliad ; of the rest only two apparently are mentioned in the accounts of the voyage of the Argo which have come down to us. What is more important however is that these accounts do mention a considerable number of Bithynian localities ^ both on the coast of the Sea of Marmara and on that of the Black Sea, as well as several peoples who do not appear in the Catalogue. We must conclude then that this hypothesis in no way accounts for the peculiarities of the Catalogue. The true explanation is not far to seek. It was well known to the ancients that the Bithynians were an intrusive Thracian people who had crossed over from Europe. Again, the Aegean coast, at all events from the Gulf of Adramyttion southwards*, was covered with Aeolic and Ionic settlements. Greek tradition unanimously held that these settlements were planted subsequently to the Trojan War, and we have no reason for sup- posing that the author of the Catalogue thought differently. At the same time we may infer from his silence that he did not claim to know what peoples had occupied these regions previously. On the other hand he did 1 Cf. Niese, Der homerische Schiffskatalog, p. 53 ff. 2 The argument that Abydos and Sestos cannot have been connected except in a irepiirXovs (cf. Niese, o/>. cit. p. 54) is one which will appeal probably to no student of early Teutonic history. ' Cf. ApoUon. Rhod. i ii64ff., II 649 ff., 720 ff., 901 ff., etc. * It is held that the Aeolic settlements in the south of the Troad were established at quite a late period, probably in the seventh century. But we are not justified in assuming that no such settlements existed previously. It is highly probable that the whole of this district was devastated by the barbarians whose remains have been found at Troy (cf. p. 295, note). 246 THE TROJAN CATALOGUE [CHAP. claim such knowledge of the previous occupants not only of the Troad but also of the Carian coast. This fact surely furnishes the means of dating the composition of the Catalogue with a fair amount of probability. It is generally agreed that the Ionic settlements were later than the Aeolic ; but probably no one will contend that the Greek occupation of Miletos and Mycale began appreciably later than the end of the tenth century. It is not necessary of course to suppose that these places were still in Carian hands when the Catalogue was composed. But at the same time the memory of their former possessors is not likely to have been perpetuated in an incidental reference like this much more than a century after they became Greek settlements. The more northern part of the list contains at all events nothing incompatible with the view that it was drawn up in the first half of the eighth century. The excavations in the Troad have certainly brought to light the fact that that district was overrun by barbarians. But there is nothing to show that this took place at the time of the Bithynian invasion, whether the two movements were connected in any way or not. The evidence on the whole seems to indicate that the barbarian occupation occurred not very long before the foundation of the Ionic colonies on the Hellespont. There is nothing very remarkable in the fact that the places mentioned in the Catalogue lie chiefly on the coast ; for such places would naturally be the most familiar to the Asiatic Greeks, who must have been a seafaring people to some extent from the beginning. Now in other parts of the Iliad we find mention of several peoples which apparently occupied the districts left blank in the Catalogue. Thus in x 429, XX 329 we hear of the Caucones, who seem to have belonged to Bithynia^. In XXI 86 f., VI 396 f. (cf. II 691, etc.) we meet with peoples called Leleges and Cilices, the positions of which are quite definitely stated. The former are said to have dwelt in the valley of the Satnioeis, in the south of the Troad, the latter somewhat further to the east, about the Gulf of Adramyttion. Again, in Od. xi Sigfif. we hear of a certain Eurypylos the son of Telephos, chief of the Ceteioi. Telephos and his son figured in the Cypria and the Little Iliad ; and Greek tradition placed their home (Teuthrania) in the region between the Gulf of Adramyttion and the Hermos. The name Kr]Teioi has been connected with that of the Hittites (Kheta) ; but without going into this question we may probably follow Strabo (xill r. 70) in tracing a reminiscence of them in the name of a stream called Ceteios, a tributary of the Caicos. The Catalogue contains no reference to any of these peoples, presumably because their names were not familiar to the author. In later times we hear no more of the Ceteioi, while Cihces are found only in Cilicia. The Leleges indeed are mentioned frequently and on both sides of the Aegean, but only as a people of the past. It is an easy and popular method of interpretation to discredit evidence for which no obvious explanation is forthcoming ; and following this method ^ Cf. especially Strabo xii 3. 5. XI] THE TROJAN CATALOGUE 247 many scholars have regarded the names under discussion as phantoms. The point against which adverse criticism has chiefly been directed is the location of the Cilices in the Gulf of Adramyttion. It is to be remembered however that we have no evidence earlier than the seventh century for the presence of a people called Cilices in the land which ultimately bore * their name. In the eighth century this people apparently dwelt to the north of the Taurus. In earlier times we hear of them, so far as I am aware, only in the ' Poem of Pentaur,' a work which celebrates the battle fought at Kadesh early in the thirteenth century, between Rameses II and the Hittites. It is there stated that the "chief of Kheta had come, having gathered together all countries from the ends of the sea to the land of Kheta, which came entire^" In addition to a number of Syrian names which occur in the accounts of earlier wars the poem contains a group of new names of peoples, consisting of Pidasa, Dardenut, Masa, Kalakisha and Luka, with two others 2. The last two of these names are almost universally identified with the Cilices and Lycians, and the first is usually connected with the name llr]haa-os or n^Sao-a, while many scholars accept the identi- fication of the second and third with the Dardanoi and Mysians respectively. If the names Pidasa and IljjSacros' are connected we are brought to the Aegean, where we find both Pedasos on the Satnioeis and Pedasa^ in Caria. Evidence to the same effect is furnished by the procession of ten warriors depicted on one of the monuments*. Of these five are of the Hittite type and two Semitic ; but the other three are of Aegean physiognomy and wear different varieties of that feather headdress which is known in Crete from much earlier times ^. When we find that this earliest reference to the Cilices associates them with a group of peoples ^ of whom some clearly belong to the west of Asia Minor, and some quite probably to the north-west corner of the peninsula, it must be admitted that we have no valid ground for discrediting the evidence of the Iliad as to their presence around the Gulf of Adramyttion. ^ Cf. Breasted, Ancient Records, Egypt, iii p. 138. '^ One of these names has been variously read as Maunna or Ariunna {Arzvena) and identified with the MT^ofes and 'IXiov of the Iliad, as well as with Oroanda and other places. The other, Keshkesh, seems to bear the same relationship to the cuneiform Kasku which the Eg. Kalakisha {Kelekesh in Breasted's orthography) bears to the cuneiform Hilakku (beside Hilak). If -ku is a suffix Kasku may possibly be con- nected with Rapes (which seems to represent an earlier Ka{s)-ar-). ^ Perhaps also called Pedasos (cf. Herod. V 121). According to Strabo (xiii i. 59) this place also belonged to the Leleges who, he says, once possessed a considerable part of Caria and Pisidia. ■• Reproduced in W. Max Miiller's Asien u. Europa, p. 361. " Cf. p. 190, note 1. * There is no question here of a national migration on the part of these peoples. They were mercenaries hired by the Hittite king, who "left not silver nor gold in his land (but) he plundered it of all his possessions and gave to every country, in order to bring them with him to battle." (Cf. Breasted, op. cii., pp. 129, note, 138.) 248 THE TROJAN CATALOGUE [CHAP. XI On the other hand it must not be overlooked that later authorities know nothing of Cilices in this region ; neither do they mention Leleges in the south of the Troad or Ceteioi anywhere. If these peoples had survived the Aeolic invasion it is difficult to believe that they could have perished subsequently without leaving some trace of their existence in Greek tradition. The Pelasgoi of Larissa in the valley of the Hermos^ appear to have been remembered in tradition, although their territories were occupied by the Aeolians — probably at quite an early date. The presumption then is that the other peoples had already been destroyed by Greek raids — as is stated in the Iliad — or else that they had been expelled or absorbed by the sur- rounding nations before the Greek colonies were fully established. In either case we shall have to carry back the poetic traditions relating to them practically to the beginning of the first millennium. We have seen that it is difficult to date the composition of the Trojan Catalogue — at all events the latter part of it — after about the middle of the eighth century. I am not aware of any valid reason for denying that the Cypria as a whole may have been composed about this time or for supposing that the Catalogue ever existed independently. It is in the preceding period — presumably between the eleventh and eighth centuries — that we must place the composition of the Iliad, though its form may not have been finally settled when the Cypria came into existence. Any more definite conclusion is rendered difficult by the unsatisfactory nature of our information regarding the Cypria. But the general impression conveyed by the epitome is that an appreciable portion of the poem was derived from incidental references in the Iliad, and that as a whole it possessed to a considerable degree the characteristics of popular poetry. If this im- pression is correct we must conclude that heroic court-poetry was in its decadence when the Cypria was composed ; and consequently we shall do well to place the flourishing period, which produced the great epics, at least a century earlier. ' Cf. Strabo xiil 3. 3 f. I see no reason for doubting the existence of traditions relating to the presence of Pelasgoi in this region. The identification of Larissa Phriconis with the Larissa of the Iliad (11 841, xvii 301) seems to be at least as likely as any of the others which have been proposed. CHAPTER XII. SUPERNATURAL ELEMENTS IN THE HOMERIC POEMS. In Chapter VI we saw that the heroic poetry of the Teutonic peoples was very largely affected by folk-tales ; that supernatural beings were frequently introduced, while ordinary human beings or animals were credited with supernatural properties — in short that the distinction between natural and supernatural was not clearly drawn. We saw further that these features were by no means confined to the later stages of heroic poetry — that on the contrary some of them were prominent even in the Anglo-Saxon poems, while the others appeared to be of equal antiquity. The same phenomena appear in Greek heroic poetry. Mythical beings and features obviously derived from folk-tales figure quite as frequently as in the Teutonic poems. Their presence is often regarded as a proof that the stories into which they enter and the persons with whom they are brought into contact are themselves products of myth or fiction. This is a question with which we shall have to deal in the following chapters. For the present it will be sufficient to quote what may be regarded as a typical expression of the attitude of more cautious scholars towards the problem of the story of Troy^: "It is fantastic to treat the siege of Troy as merely a solar myth — to explain the abduction of Helen by Paris as the extinction of the sunlight in the West, and Troy as the region of the dawn beset and possessed by the sunrise. It is equally fantastic, and more illogical, to follow the rationalising method — to deduct the supernatural element, and claim the whole residuum as historical ^ Jebb, Introduction to Homer, p. 147. 250 THE SUPERNATURAL IN THE HOMERIC POEMS [CHAP. fact. Homer says that Achilles slew Hector with the aid of Athene. We are not entitled to omit Athene, and still to affirm that Achilles slew Hector." It may be observed that the goas are introduced in the Homeric poems in many different ways. The incident just cited — where Athene takes the form of Deiphobos (II. xxii 226 ff.) — belongs to one of the commonest types, and one which requires comparatively little imagination, if we are prepared to grant the existence of a belief that the gods were capable of disguising themselves in human form. Sometimes again deities render themselves visible only to certain individuals out of a crowd, as in II. I 194 ff, where Athene intervenes in order to stay Achilles from drawing his sword upon Agamemnon. Another type is the disguise of gods as birds, as when we find Athene and Apollo sitting upon an oak in the form of vultures, before the combat of Hector and Aias {ib. Vli 58 ff.). On other occasions birds are sent by a god as a sign of his favour or protection, as in II. XXIV 315 ff- and many other passages. Somewhat akin to this type is the dream sent by Zeus to Agamemnon in II. II 5 ff It has been remarked that in such cases as these a sceptical person might have accounted for everything that passed without reference to any intervention on the part of a deity. But there are a number of other cases where the action is affected by gods in ways which could not be accounted for on any rationalistic hypothesis. We may refer to II. Ill 380 ff., where Aphrodite snatches Paris away from Menelaos and conveys him to his own house in Troy. Or again to several passages in the Diomedeia where deities show themselves almost or quite without disguise and even take an active part in the fighting. A still greater amount of imagination perhaps is required for the scenes which depict the quarrels and amusements of the gods in Olympos, and their schemes for helping or destroying the combatants. It has been held that all such passages as these belong to a later period than those of the less imaginative types described above. In Teutonic stories of the Heroic Age, as we have already seen, very few notices relating to the gods have been preserved. XII] THE SUPERNATURAL IN THE HOMERIC POEMS 25 1 The appearances of Othin in Volsunga Saga (cf. p. 1 14) may be compared with those of Apollo in the Iliad or Athene in the Odyssey ; for his divinity is not recognised at once, though he does not take the form of a person known to his favourites. Again, though I know no exact Homeric parallel to the incident in Sigmundr's last battle, when the hero's sword is shattered at the touch of Othin's javelin, the idea is in complete harmony with several passages in the Diomedeia (e.g. V i2gf{., 438 ff., VI 128 ff., 306 f). It is true that we do not know whether these incidents in Volsunga Saga are based on old tradition or not. But a good parallel for the last of the Homeric types is furnished by the Langobardic story of Wodan and Fria and the victory granted by them to the Winniles (cf p. 115). The similarity between this story and the incident related of Zeus and Hera in II. XIV 153 — 353 gives us some ground for suspecting that the heathen poetry of the Teutonic Heroic Age may have possessed decided ' Homeric ' characteristics in its treatment of the gods. As the case stands however we shall have to take our illustrations from stories of the Viking Age, A somewhat curious parallel to the incident of Athene and Hector in II. XXII 226 ff is to be found in the story of Haraldr Hilditonn, as told in Saxo's History, pp. 255, 263. Haraldr had a con- fidential servant named Bruno, whom he employed to drive his chariot and to carry messages to his nephew Ringo (SigurSr Hringr). This man eventually was drowned ; but Othin took his place and form, and exerted himself to sow discord between the two kings. It was not until the battle at Bravalla had begun that Haraldr had any suspicion of the treachery which had been played upon him. Then suddenly recognising the identity of his charioteer he begged him to grant him victory. But Othin threw him out of the chariot and slew him. The story of Haraldr Hilditonn refers, it is true, to times for which we have no historical records. But a still more graphic story of intervention on the part of divine beings occurs in con- nection with a well-known event — the expedition made against Norway by the Jomsvikingar. In the latter part of the tenth century a number of Scandinavian adventurers had established 252 THE SUPERNATURAL IN THE HOMERIC POEMS [CHAP, and fortified a settlement at Jomsborg on the island of Wollin at the mouth of the Oder. About the year 994 their leaders, Sigvaldi and Bui, made a vow to attack Haakon, earl of Lade, who then ruled Norway. The earl was taken by surprise and had not been able to muster all his forces when he encountered the hostile fleet at Hiorungavigr, near the mouth of the Romsdal Fjord. The battle at first went against him ; and, according to the story, he took advantage of a respite in the fighting to retire to one of the islands and pray to Th6rgert5r HolgabruSr. He was not able to obtain her assistance until he had sacrificed to her his youngest son. When he resumed the fight, the weather, which had been hot, underwent a complete change. A snow-storm came from the north and beat in the faces of the pirates, so that they were numbed with the cold and could neither move nor see. But worse was to come. " It is said that HivarSr, one of Bui's companions, was the first to see Th6rgert5r in Haakon's fleet ; but soon she was seen by many, both those who had second sight and those who had not. When the snow abated a little they saw also that arrows were flying, as it seemed, from every one of the demon's fingers, and each arrow brought about a man's death." Then they tell Sigvaldi, who says : " It seems to me that we have got to fight to-day not against men, but against the worst of devils." Still he continues the fight. When Haakon saw that the snow was abating he cried with all his might to ThorgerSr and her sister Irpa, reminding them how much he had given up to them in sacrificing his son. Then the storm began again, and soon HavartJr saw two female figures on Haakon's ship, both acting as the one had done previously. Then Sigvaldi said that he would now take to flight and that all his men were to do likewise, for they had made no vow to fight against devils ^ Here also we may cite a story connected with another historical event, which took place about ten years before the battle of Hiorungavagr. Eric the Victorious, king of Sweden, expelled his nephew Styrbiorn from the kingdom ; and the latter 1 Jomsvikinga Saga, cap. 44 (Fornmanna Sogur, XI p. 136 ff.) ; Flatey-iarbok, I 191 f. Snorri gives a different account of the battle in the Heimskringla (Olafs S. Tryggv. 43 ff.) ; but he was acquainted with at least part of the story given above. XII] THE SUPERNATURAL IN THE HOMERIC POEMS 253 invaded the country with the help of the Danish king Harold Blue-tooth. On the eve of the battle Eric went into Othin's temple and in order to obtain victory promised to give himself up dead at the end of ten years. Soon afterwards he saw a big man with a long hood, who put a cane into his hand and told him to throw it over Styrbiorn's army saying : " You all belong to Othin," When he threw the cane it seemed to turn into a javelin and brought blindness upon Styrbiorn and all his host. On the same occasion Thor was seen in Styrbiorn's camp^ Such cases are by no means isolated. Olafr Tryggvason, who reigned over Norway from 995 to looo, is said to have been visited both by Othin and Thorl Of dreams perhaps the most interesting case is a story told of an Icelander named Glumr, a contemporary of Earl Haakon of Lade. A certain Thorkell possessed an estate which he was compelled to sell to Glumr. Before leaving he went to Prey's temple, sacrificed an ox and prayed that Glumr likewise might be forced to give up the estate. This actually came to pass. But before he left, Glumr dreamed that he saw a great crowd on the river banks coming to see Frey, who was seated on a chair. In his dream he asked who they were. They replied that they were his departed relatives and that they were praying Frey that he (Glumr) should not be driven from his estate. But it was of no use ; Frey answered curtly and angrily, remembering the ox which Thorkell had given him^. For the action of Homeric deities in sending birds as a mark of favour or omen of success a good parallel is to be found in another incident in the life of Earl Haakon of Lade. When Jutland was invaded by the Emperor Otto H in 974 Harold Blue-tooth summoned Haakon to his assistance. After the campaign the Danish king adopted Christianity and compelled Haakon to do likewise. But the latter set off with his fleet as soon as possible and, landing on the coast of Ostergotland, proceeded to offer a great sacrifice. Thereupon there came two ^ Styrbiarnar Thattr, cap. 2 (Fornm. Sog., v p. 250). ^ Olafs S. Tryggv. A (Heimskr.), cap. 71; Olafs S. Tryggv. B, cap. 213 (Fornm. Sog., II p. 182 f.). ■* Vfga-Glums Saga, cap. 9, 26. 254 THE SUPERNATURAL IN THE HOMERIC POEMS [CHAP. ravens flying by and screaming loud. The earl interpreted this as a sign that Othin had accepted the sacrifice and that he would have a favourable time for battle^. The importance of the last case is enhanced by the fact that it is derived from a contemporary poem, the Vellekla (cf p. i6), in which the ravens are mentioned. In the other cases given above no such early authority is extant, and our texts themselves are separated by a period of from two to three centuries from the events which they relate. But there is no reason for doubting that the stories had long been in existence. It is questionable indeed whether the account of the battle of Hiorungavagr could have been invented after all recollection of Haakon and his religious observances had died away. The earl's devotion to the worship of ThorgertJr HolgabruSr is known from other sources ; but she seems not to have been a generally recognised member of the Northern Pantheon. However that may be, no doubt can be entertained with regard to the poem Hdkonarmal, which deals with the death of King Haakon I at the battle of Fitje in 961 (cf p. 15). The author, Eyvindr Skaldaspillir, was himself present at the battle. The poem relates how Gondul and Skogul were sent by Othin to select a prince of Yngvi's line, who should go and dwell with him in Valholl. Then, after a short account of the battle, we are told that " the princes sat with their swords drawn, with scarred shields and mail-coats pierced ; in no cheerful mood was the host which had to make its way to Valholl. Then said Gondul, as she leaned upon her spear : ' Now will the forces of the gods be increased, since they have summoned Haakon to their abodes with a great host' The prince heard what the noble Valkyries were saying. Thoughtful was their mien, as they sat on their steeds, with helmets upon their heads and holding their shields before them. ' Why hast thou thus decided the battle, Skogul ? Surely we have deserved success from the gods.' ' We have brought it about that thou hast won the day and that thy foes have fled. Now,' said the mighty Skogul, ' we must ride to the green homes of the gods, to tell Othin that a monarch is coming ' Olafs S. Tryggv. (Heimskr.), cap. 27 f. XII] THE SUPERNATURAL IN THE HOMERIC POEMS 255 to enter his presence.' " Then the scene changes to Valholl ; and Othin sends HermoSr and Bragi out to meet the king and bid him welcomed Such poems as H^konarmdl and Eiriksmdl must be regarded as products of vivid poetic imagination. They are clearly in the nature of conscious fiction, though it should not be assumed that the pictures of the gods and their abode which they present were conceptions altogether unreal to the poets' audiences. I suppose that ultimately this type of composition is derived from visions or dreams, such as the story of Glumr given above. For, though the latter in its present form dates from a period at least two centuries later than Hakonarmdl — both referring to persons who lived more or less about the same time — it will probably be agreed that the conception there is far more primitive. Now we have good evidence that visions which took the form of visits to the home of the gods did really obtain credence in the Viking Age. When St Ansgar visited Sweden for the second time, not long after 850, he found that the success of his mission was seriously endangered. A man had come to Birca (Biorko, on the Malar), where the king, Olaf, was residing, and stated that he had been present at an assembly of the gods, who had sent him to deliver a message to the king and nation. This was to the effect that the gods had long been gracious to the Swedes and had preserved their land in peace and prosperity. Yet now the Swedes were abandoning their accustomed sacrifices and introducing a strange god. If they wished to retain their favour the sacrifices must be resumed on a larger scale, and the new god must be refused admittance. " But if you desire to have more gods, and we are not sufficient for you, we unanimously enrol in our body Eric who was formerly your king, so that he shall be counted among the gods." This story created a profound impression among the inhabitants. " They founded a temple in honour of the above-mentioned king, who had ^ This latter part of the poem is copied from Eiriksm^l (cf. p. 15), in which Othin sends out Sigmundr and Sinfiotli to meet Eirikr. We do not know either the date or the author of Eiriksmil ; but it would seem from the Saxon Chronicle that Eirfkr was still alive about the year 954. 256 THE SUPERNATURAL IN THE HOMERIC POEMS [CHAP. long been dead, and began to offer prayers and sacrifices to him as a god^." The Life of St Ansgar, from which this story is taken, is practically a contemporary authority. It was written by St Rembert, one of Ansgar's disciples, who succeeded him as Archbishop of Hamburg in 865 and died in 888. In face of such evidence we have no reason for doubting that stories such as that of Glumr would readily obtain credence. Indeed it seems scarcely impossible that the doings of ThorgerSr HolgabruSr at Hiorungavagr may have been believed by persons who were alive at the time. Yet credulity was no special characteristic of the Northern peoples. We learn from inscriptions that Asclepios was in the habit of showing himself to pilgrims in his temple at Epidauros. Still more striking evidence is furnished by Herodotus' story (I 60) that Peisistratos recovered the tyranny at Athens by dressing up a woman to personate Athene and accompanying her in a chariot to the city. Herodotus himself remarks that this took place at a time when the Greek race had long been distinguished from the rest of mankind by its superior sagacity and freedom from silly credulity — and in a state too which was held to be intellectually supreme among the Greeks. Whatever doubt may be entertained as to the truth of the story it is significant enough that Herodotus should record it, apparently without any hesitation, in less than a century after Peisistratos' death. Some four centuries earlier men may well have been ready to hear that the gods took an active part in the battles of their fathers or grandfathers, while the latter themselves may have been quite as ready to attribute their success or failure to the disguised agency of the same powers. It is clear at all events that the Scandinavian evidence fails to provide any justification for the view that poems which introduce the gods must date from times far removed from the events which they claim to commemorate. In the contemporary ^ Porro, si etiam plures deos habere desideratis, et nobis non sufficimus, Ericum, quondam regem uestrum, nos unanimes in collegium nostrum asciscimus, ut sit unus de numero deorum Nam et templum in honore supradicti regis dudum defuncti statuerunt, et ipsi tanquam deo uota et sacrificia offerre coeperunt. Rem- bertus, Vita S. Anscharii, cap. 23. XIl] THE SUPERNATURAL IN THE HOMERIC POEMS 257 Hakonarmdl we find two of the most advanced Homeric types. First, we have deities participating without disguise in battle ; then a change of scene carries us to the actual home of the gods. The second scene in Hakonarmdl — which is likewise the scene of Eiriksmal — may be compared both with the various 'Olympic' episodes and also with the two Nekyiai. For, since Othin is a god of the dead, his abode corresponds in a sense both to Olympos and the home of Hades. On the whole perhaps the nearest affinities of the two Norwegian poems are with the second Nekyia (Od. XXIV i — 204), which is commonly regarded as one of the latest portions of the Odyssey. Lastly, the Scandinavian evidence gives no support to the belief that the more imaginative types of divine intervention necessarily belong to a later date than the others. In principle of course it may be admitted that they are less primitive. But in Old Norse literature it so happens that they occur in both earlier and more nearly contemporary works. The explanation lies doubtless in the fact that the theological apparatus of Norse poetry was fully developed before the time of our earliest authorities. So far as I can see, there is no good reason for denying that the same remark holds good for the Homeric poems. But if so it is futile to use evidence of this kind as a criterion for determining the date of the various portions. Monsters and theriomorphic demons are by no means unknown in Greek heroic stories, though in the poems which have come down to us they figure prominently only in episodes dealing with past events. The nearest Homeric analogies to Beowulf's adventures are perhaps to be found in the stories of Bellerophon and Meleagros^ (II. VI 178 fif. ; IX 538 ffi). It may be that these stories and others, such as that of Perseus, them- selves once formed the main themes of heroic poems, and that the backward position which they ultimately came to occupy, as compared with stories of anthropomorphic deities, is due to the growth of poetic art and humanistic tendencies. But on the * The first stage in the growth of such a story as this may be illustrated from the message of the Mysians given by Herodotus, i 36. The development which it may ultimately attain can be seen from the story of Kilhwch and Olwen. 258 THE SUPERNATURAL IN THE HOMERIC POEMS [CHAP. other hand we have to rernember that heroic poetry is always liable to the intrusion of folk-tales, in which adventures with monsters form one of the favourite themes. The chief store-house of folk-tales in the Homeric poems is the narrative of his adventures given by Odysseus to the Phaeacians (Od. ix — xil). This narrative contains a consider- able number of incidents, of which ten may be regarded as more or less distinct : (i) the encounter with the Cicones, (ii) the visit to the land of the Lotus-eaters, (iii) the adventure with Polyphemos, (iv) the two visits to Aiolos, (v) the disaster in the land of the Laistrygones, (vi) the two visits to Circe, (vii) the journey to the home of Hades, (viii) the singing of the Sirens, (ix) the adventure with Scylla and Charybdis, (x) the slaughter of the cattle of Helios. The first of these incidents bears no obvious traces of de- rivation or influence from a folk-tale ; but it is the only one of the series of which this can be stated with any confidence. The adventures with the Lotus-eaters and the Laistrygones should perhaps be regarded rather as travellers' stories — founded possibly on actual experience of foreign peoples — yet the latter at least contains certain distorted features which may fairly bring it within our category. As to the origin of the adventure with Polyphemos there can be little doubt. It appears to be found with slight variants in many different parts of the worlds The last incident of the series is perhaps the one least widely known; but a parallel may be cited from one of Saxo's stories (p. 286 f.). A certain Danish king named Gormo^ was an ardent explorer. Above all he desired to visit the abode of Geruthus (GeirroSr), which lay beyond the ocean in a land of perpetual darkness. Taking with him as guide an experienced traveller, named Thorkillus, he set sail with three ships and made his way beyond Halogaland (the north of Norway). There, having lost its way in a storm, the expedition came to be ^ Cf. Macculloch, The Childhood of Fiction, p. 279 ff. ^ The historical connections of this story are somewhat obscure. But this is immaterial for our purpose, as the part with which we are dealing is clearly derived from folk-tales. XII] THE SUPERNATURAL IN THE HOMERIC POEMS 259 in want of food. Eventually they arrived at an island which contained herds of extremely tame cattle. Against the advice of Thorkillus the mariners slaughtered a large number of these. The following night they were attacked by monsters, one of whom declared that they would not be allowed to sail away until they made compensation for the losses they had inflicted on the herd of the gods. In order to save themselves they had to give up one man from each ship. It will be seen that this incident bears a general resemblance to the slaughtering of the cattle of Helios, and we need scarcely hesitate to regard both stories as variant forms of a folk-tale. As to its origin we are not altogether without evidence in the Northern case. In Alcuin's Vita Willebrordi, I cap. 10, it is stated that a certain island (now Heligoland) was entirely sacred to a god named Fosite. So great was the sanctity with which it was regarded that no one ventured to touch any of the animals which grazed upon the island. The violation of the sanctuary, in this and other respects, cost one of St Willebrord's com- panions his life. Hence there is no need to doubt that a basis of fact underlies the stories of islands in which animal life was held sacred — ^just as in holy woods throughout the north of Europe. It is scarcely impossible that similar island sanctuaries may once have been known in the Mediterranean. The subsequent course of the story has a certain affinity with that of Circe. After leaving the island Gormo and his men sailed in safety to the farther part of Permland, where they were met by a giant named Guthmundus\ the brother of Geruthus, who invited them to his house. Thorkillus strictly enjoined his companions to abstain from all food and drink offered them, even from the fruits which grew in the garden, and to avoid contact with members of the household. Those who yielded to temptation, as a few eventually did, would have to spend the rest of their lives among monsters. There is no actual transformation as in the story of Circe; but this in itself is a widely known incident in folk-tales. When the travellers at length reach the abode of Geruthus • Gu^mundr of Glsesisvellir is a well-known figure in the unhistorical parts of sagas. 26o THE SUPERNATURAL IN THE HOMERIC POEMS [CHAP. the scene, though horrible in every way, seems to be a variety of the Enchanted Castle rather than a parallel to the home of Hades^ We have seen above that the poems Hdkonarmdl and Eiriksmal may in a sense be compared with the two Nekyiai ; for Valholl is the abode not only of the chief god but also of the spirits of fallen warriors. But here we have to deal with elaborate conceptions of court poetry which are further removed from the spirit of the true folk-tale than either of the passages in the Odyssey'^. A better parallel to the first Nekyia is perhaps to be found in another of Saxo's stories (p. 31). Once upon a time, when King Hadingus was feasting, there appeared to him a woman who was carrying hemlocks. She wrapped him in her mantle and took him with her underground in order to show him where the hemlocks grew. On the way they passed through a dark cloud and then along a well-worn path, where they saw many men richly attired. After viewing the sunny regions where the hemlocks grew, they crossed a rapid river and then saw two armies engaged in desperate conflict. The woman told Hadingus that these were men who had been slain by the sword and continually rehearsed the manner of their death. They are obviously to be connected with the emJieriar of Old Norse poetry — the slain warriors who dwell in Valholl and spend their days in combat — though possibly this passage represents a more primitive form of the idea It is to be observed that Saxo himself explicitly interprets the story as a visit to the region of the dead. Stories of this kind are to be found in many parts of the world — among peoples as widely apart as the Algonquins, the Zulus and the Maoris^ There can be little doubt that to a large extent the first Nekyia belongs to the same category. At the same time of course I do not mean to imply that it is wholly to be regarded as a folk-tale. In the interview with Agamemnon and his companions (vv. 385 — 564) we find ourselves in much the same world of ideas as is presented to us in Eiriksmal and ^ A better parallel is perhaps furnished by Thorkillus' subsequent visit to the abode of Ugarthilocus (p. 292 ff.). The description of this place recalls that of Nastrond in Voluspa 39, Gylfaginning, cap. 52. ^ The home of Hades resembles the abode of Hel rather than Valholl. To this also we have a visit (by the god HermotSr) in Gylf. 49. * Cf. Tylor, Primitive Culture^, i p. 346, \\ p. 50 ff. XIl] THE SUPERNATURAL IN THE HOMERIC POEMS 261 Hakonarmal. Again, it is to be remembered that the object of Odysseus' journey was to consult the spirit of Teiresias, and this is perhaps the original kernel of the story. Such an idea however may be derived from ancient religious observances rather than from a folk-tale. Herodotus (V 92) records that Periandros, tyrant of Corinth, about the close of the sixth century, sent an embassy to the oracle of the dead {veKvo- fiavrrjiov) on the river Acheron in Thesprotis, in order to consult the spirit of his wife Melissa. After making all allowance for antiquarian and etymological speculation^ it seems probable that this oracle did influence the conceptions of the home of the dead current in Greek poetry. It would appear then that in the composition of the first Nekyia we have to take account of the influence of at least three different elements — court poetry, folk-tale and religious (necromantic) observances. If we are right in supposing that Aeolis was the true home of all Homeric poetry, the absence of any precise geographical indications is easily accounted for. During the centuries which intervened between the end of the Heroic Age and the beginning of the historical period there is extremely little evidence, whether traditional or archaeological, for communication with distant lands; and it is likely enough that at that time Thesprotis was as unfamiliar as Egypt to the inhabitants of Aeolis. Few scholars will dispute that the geographical indications throughout the story of Odysseus' wanderings are both vague and contradictory. Sometimes he appears to be in the west ; sometimes again he is following the track of the Argo — presumably in the Black Sea. That is after all the kind of confusion which might reasonably be expected from poets who were dealing with traditions of voyages made long before in regions now altogether forgotten. The ascription of supernatural properties to men or animals is not a very striking feature in Homeric poetry — unless we ^ As seen (e.g.) in the application of the name KuKvrbs to a tributary of the Acheron. The presumption is that this name was originally a creation of poetic fancy, just as much as Ylvpi(p\iyid(j)v. The diffusion of the names Acheron and Acherusia in other regions (Italy, the Black Sea, etc.) is doubtless due to the influence of poetry or tradition. 262 THE SUPERNATURAL IN THE HOMERIC POEMS [CHAP. XII include under this head stories of exaggerated prowess. As an example we may cite II. xix 404 ff., where one of Achilles' horses speaks and prophesies his master's death. Incidents such as the flame on the same hero's head in II. XVIII 205 fif. and the changes in Odysseus' appearance (Od. XlII 429 fif., etc.) are attributed to the direct action of deities. On the other hand exaggeration is common and often carried out systematically. Among such cases we must include the feats of valour performed by some of the combatants, and also presumably the numbers of the forces stated in the catalogues, if we admit that the story of the siege of Troy has any historical foundation. On the whole it appears that those elements in the Homeric poems which may quite safely be derived from myth or folk-tale resemble the corresponding elements in Teutonic heroic poetry very closely. We may perhaps doubt whether the gods ever figured so conspicuously in Teutonic poetry as they do in the Iliad and Odyssey ; but the difference between the two cases is one of degree only. In the use made of folk-tales the difference is very slight. It remains for us now to consider whether the remaining elements in the poems — their main groundwork in fact — should be regarded as of similar origin in both cases. CHAPTER XIII. MYTH IN THE HOMERIC POEMS. It is commonly held that history, myth and fiction have all contributed to the formation of the Greek heroic stories ; but opinions differ widely as to the relative importance to be attributed to the three elements. Among modern scholars the general tendency has been to assign the chief weight to myth. By many indeed the heroes of the Trojan War are believed to be as mythical in origin as the gods themselves. One conclusion may safely be drawn from the Northern evidence discussed in the last chapter : we must definitely dis- miss the argument that the Homeric heroes cannot have been men of flesh and blood because they are brought into contact with the gods. No one will be so hardy as to suggest that King Haakon or his namesake, the famous earl of Lade, were products of myth or poetic imagination. Yet Gondul is as much responsible for the death of King Haakon as Athene is for that of Hector. There is certainly this difference between the two cases, that we have no historical evidence for the existence of the Homeric heroes. But the fact that deities participate in their destruction does not in itself prove that they are themselves products of myth or fiction. There was a time, not so very long ago, when most of the characters of the Greek Heroic Age were believed to owe their origin to nature-myth — personifications of light, darkness and so forth. At the present time however it is only in some few cases that this view is generally maintained. Its chief strong- hold is the case of Achilles ; and here we are invariably referred for proof to the story of SigurtJr. The two characters have of 264 MYTH IN THE HOMERIC POEMS [CHAP. course a good deal in common. Both are more or less idealised types of youthful strength and valour, and both die prematurely. But it would be pure folly to regard these features as in them- selves proofs of mythical origin. In order to prove this it is necessary to point to features which can only be mythical, and to show that such features formed an original element in the stories. Now we have seen (p. 140 ff.) that the current explanation with regard to SigurSr is open to the most serious — in my opinion fatal — objections. On the other hand there certainly was a tendency for myth to grow up in later times round this hero. As an instance we may take his invulnerability, a feature which is peculiar to the German version of the story. Achilles possesses the same characteristic — but not in the Iliad or Odyssey. It is as much unknown in the Homeric account of Achilles as in the Norse account of SigurSr. Indeed the only essentially mythical feature which the poems themselves record in the case of Achilles — and it is by no means peculiar to his case — is that he is the son of a deity\ But divine descent was claimed also by many Teutonic princes, though the heroes of our stories are usually separated from their divine ancestors by two or three generations-. Whatever may be the explanation of this phenomenon it is doubtless to be connected with the stories of conjugal relations between human and divine beings which we find both in Greece and in northern Europe. This is a subject to which we shall have to return in a later chapter. Above all, however, we have to take account of the influence of folk-tales^ and popular beliefs, which, as we have seen from 1 Achilles himself was worshipped as a deity in certain localities ; and the same is true of some other heroes. We may refer to the story of St Ansgar, quoted above (p. 255 f.). With such cases as that of ' Zeus Agamemnon ' we shall have to deal later. 2 According to Volsunga Saga, cap. 2, Sigmundr, the father of Sigur'Sr, had a divine mother (cf. p. 114). ^ Some resemblance to the case of Peleus and Thetis is shown by a story in Hrolfs S. Kraka, cap. 15, where an elf-woman bears a daughter to Helgi, the father of Hrolfr Kraki. Such incidents are not uncommon in folk-tales. We may note especially those cases in which the supernatural bride is a mermaid, perhaps re- presenting the Swan-maiden of earlier times. Thetis has a good deal in common with the latter class of beings. XIIl] MYTH IN THE HOMERIC POEMS 265 the Teutonic evidence, may make itself felt even in the descrip- tion of very recent events. The story of the abduction of Helen is another case for which many scholars still claim a mythical origin. It is perfectly true that stories of (e.g.) the abduction of the sun or the incon- tinence of the moon' do occur, though examples of this type are by no means so common or widespread as many writers have assumed ^ But what is apt to be overlooked is that these stories arise from a personification of the sun or moon, and that it is in consequence of this personification that the heavenly bodies are believed to be exposed to perils and passions such as affect human beings. It is surely nothing less than an inversion of the natural order of things to suppose that the numerous class of folk-tales which deal with the abduction of a girl or wife originated in the type — a comparatively rare type — in which this motif is applied to the sun. There can be no reasonable doubt that the prevalence of such folk-tales is due to the in- numerable occurrences of abduction in real life. But the theory we are discussing involves not merely the personification of heavenly bodies and natural phenomena but their complete anthropomorphisation^ — a very doubtful process in the best of cases — whereas the story which it seeks to explain bears no trace even of derivation from a folk-tale. In other words we are asked to assume a most complex and precarious hypothesis ' Cf. Aeneas Sylvius, Hist, de Eur., cap. 26, and the first Daina in Schleicher's Handb. d. litau. Sprache. Night and day or dawn are also frequently personified — the last especially where, as in Greece, the sun is regarded as a male. But none of these lend themselves so readily as the sun and moon to the development of mythical stories. The personification of light, darkness, etc. in the abstract seems to belong to a much more advanced stage of thought. * Eclipse-myths (usually of a simple character) are widespread and fairly common. A probable example is to be found in Gylfaginning, cap. 12 (cf. also Tylor, Primitive Culture^, I p. 328 ff.). But a good deal of scepticism is justifiable in regard to the interpretation of stories which are supposed to have originated in myths of sunrise and sunset. This remark applies even to those Polynesian and Red Indian stories which are commonly regarded as among the best examples of their class. ^ The personification of the sun and the dawn in the Homeric poems is very similar to what we find in the north of Europe, e.g. in Gylfaginning, cap. ro f., and the first four Dainos in Schleicher's Handbuch. The most important difference is that the Dawn-goddess, like other deities, has sexual relations with mortals (see the Addenda). But her true character is not for a moment forgotten. 266 MYTH IN THE HOMERIC POEMS [CHAP. in order to account for a story for which parallels are to be found very frequently in almost all stages of human society. . Of course I do not mean to say that the story of Helen is entirely devoid of mythical elements. On the contrary, it is a most instructive example of the growth of myth, and as such it furnishes an interesting parallel to the history of similar stories in the north of Europe. In the Iliad Helen possesses no mythical characteristics, except that she is the offspring of a divine father. In the Cypria she had apparently also acquired a divine mother (Nemesis). By the seventh century we find her figuring in quite a different story of abduction — a story which seems to have been treated by Alcman and Stesichoros, as well as on the ' Chest of Cypselos.' This time she is carried off by Theseus, with the help of Peirithoos, and rescued by her brothers, the Dioscoroi^ Somewhat later we find a new version of the story of her abduction by Paris. Now it is said to be only her elBcoXov which is carried off by Paris ; Helen herself is taken by Hermes to Egyptl There seems to be little reason for doubting that the etBwXov was a deliberate invention of Stesichoros, though in other respects this version of the story may well have been influenced by the Egyptian version, recorded by Herodotus (II 1 12 ff). The latter again comes in all probability from Greek settlers in Egypt, who connected the narrative of Helen's sojourn in Egypt, related in the Odyssey (IV 125 ff., 351 ff), with a cult which they found existing in that country. This version of the story then should' perhaps be regarded as a product of fiction rather than myth. The other story however — that of Theseus, Peirithoos and the Dioscoroi — is doubtless of popular origin. It is important to notice that both these pairs of heroes are 1 There is probably an allusion to this story in II. Ill 144, although a different explanation is quoted by Plutarch (Theseus, cap. 34) from Istros. But the verse in question was condemned by some ancient, as well as modern, scholars. It is 'inorganic' (cf. Od. Ii 331) and due in all probability to the same process as II. II 831 ff. (cf. p. 244, note). ^ This list is by no means exhaustive. We may mention also the story of Helen and Achilles in the 'White Isle' (cf. Pausanias in 19. 11) and that of Ariston's wife, related by Herodotus (vi 61). The Rhodian story (cf. Paus. Ill 19. 10) is obviously due in part to the influence of the Homeric poems ; but it is at least questionable whether this 'EXivr) AevSp^Tis was originally identical with the other Helen. XIIl] MYTH IN THE HOMERIC POEMS 267 connected with other stories of abduction \ Moreover in both cases these stories have certain elements in common with that of Persephone. There can be Httle doubt therefore that we have to deal with a folk-tale. The introduction of Helen into the story may be due partly to her kinship with the Dioscoroi and partly to the influence of the story of her abduction by Paris. At the present time it appears to be the more general opinion that the Homeric heroes originated mainly not in personifications of natural phenomena, but in tribal divinities or personified conceptions of peoples (' hypostasierte Volks- individualitaten '). Now we have seen (p. 131 f) that in Teutonic heroic poetry we occasionally meet with the mythical eponymous ancestors of families, though such persons are referred to the past and not introduced into the main action of the stories. Similar characters are to be found in the Homeric poems. Perhaps the best example occurs in a speech of Aineias (II. XX 200 fif.), where the names AdpSavo^, Tpw? and 'IX09 are included in the hero's genealogy. The Cadmos of Od. V 333 is probably to be regarded, in some sense or other, as the eponymous an- cestor of the Cadmeioi, though he is not mentioned in connection with Thebes. Again, in Od. xvii 207 we have a reference to eponymous heroes of places, Ithacos and Neritos. They are perhaps creatures of the poet's own imagination, i.e. fictitious rather than mythical beings ; but it is probable that they were modelled upon existing types. Other examples of both types may be found elsewhere in the poems^. Yet it cannot be said that they are common. In Greece, as in northern Europe, the true home of eponymous ancestors (Hellen, Doros, Achaios, etc.^) is to be found in post-heroic, or at least non-heroic, literature. In recent years however several scholars have put forward the theory that the characters who figure in the main action of ^ Attention should be paid not only to the case of the Leucippides but also to the story of Phormion (Paus. Ill 16. 3). ^ In 11. II 828 ff. (if the name "ASpijo-ros is taken from vi 37 ff. or XVI 694) we have apparently the case of an already existing character being turned to account as an eponymous hero. ^ Alo\l8r]s occurs occasionally as a patronymic for individuals. The * keeper of the winds ' seems to have no connection with these characters. 268 MYTH IN THE HOMERIC POEMS [CHAP. the Iliad are tribal heroes in disguise. For a simple example of this theory we may refer to the interpretation put upon II. V 43 f(., where the Cretan leader Idomeneus is represented as slaying a man named Phaistos (ato-T09). Now there was in Crete a well-known city called Phaistos (4>ato-Td?). According to Prof E. Bethe (^Neiie Jahrbilcher, Vll 669) it cannot be dis- puted that the man Phaistos is the ' eponym ' of the city and that we have here the remains of an ancient Cretan heroic lay. But the origin of the man is stated explicitly enough in the poem {/.c): he is the son of Boros the Maeonian and had come from a place called Tarne. Before we can assume that he was the ' eponym' of a Cretan city we must surely ask how he came to be represented as a Maeonian (Lydian). Is it inconceivable that a name identical with that of a city should be borne by anyone except the eponymous hero of the city .'' This is not the only case of the kind which has been brought forward. In II. V 706 we hear of an Aetolian named Trechos slain by Hector and in II. XX 455 of a Trojan named Dryops slain by Achilles. Here we are said to have ' eponyms ' of Trachis and the Dryopes. In England during the centuries immediately following the Heroic Age we find mention in historical documents of princes or ecclesiastics called Walh, Cumbra, Seaxa, Dene, Fronca, etc. Are we to suppose that these persons are the eponymous heroes of the Welsh or Cymry, the Saxons, Danes and Franks ? But national names of this type seem to have been just as frequently used by the Greeks, at least in historical times. We may mention Achaios of Eretria, Ion of Chios and Dorieus the brother of Leonidas. Is there any reason for denying their use in earlier times^? The evidence of these names has been brought forward in support of a far-reaching theory — that the conflicts which we find described in the Iliad are echoes of tribal struggles which once took place in Greece, and that the warriors, Trojans as well as Greeks, are in reality mythical heroes in whom the various ^ In the Homeric poems it is in the case of minor characters among the Trojans and their allies that names of this type are most common. In this case the use of such names may be accounted for with considerable probability under the head of fiction (cf. p. 300, note). I XIIl] MYTH IN THE HOMERIC POEMS 269 contending tribes have become personified. If this theory is sound it will be obvious that the resemblance between Greek and Teutonic heroic poetry must be merely superficial — that the two groups of poems spring from essentially different sources. It will be well then to examine somewhat carefully the evidence on which the theory is based. The first argument in its favour is derived from a story quoted by Plutarch {Theseus, cap. 34) from Istros, a writer of the third century, to the effect that Alexandres (Paris) was overcome by Achilles and Patroclos on the banks of the Spercheios. In confirmation of this story it is pointed out that the warriors with whom Paris fights in the Iliad mostly belong to Thessaly, while his sister Alexandra (Cassandra) was worshipped by the Locrians. Another argument rests on a story derived from the Little Iliad, that Andromache was brought to Pharsalos after the fall of Troy. The inference that she belonged originally to this region is supported by the proposed identification of Thebe Hypoplacie, her home in the Iliad, with the Phthiotic Thebes, to the east of Pharsalos. Yet a further argument relates to Hector. It is noted that he was worshipped as a hero at Thebes in Boeotia, and that most of the persons associated with him, either as friends or foes, are connected with Boeotia, Thessaly and the intervening districts. In Prof. Bethe's words " Hector's tracks lead from southern Thessaly, through Phocis and Boeotia, to the Cadmean Thebes." " In other words Hector, or rather the tribe which honoured Hector as their hero, migrated by this road. More accurately, the tribe gradually, in how many centuries none can tell, moved in a south-easterly direction, driven by a pressure which was no doubt exerted by the Aeolic tribe represented in the Epos by AchillesV Now it is manifest that the argument derived from Istros' story can have validity only if it can be shown that there is reason for believing it to be based on genuine native tradition, independent of the Homeric poems. For everyone who has studied the history of Teutonic heroic poetry knows that in the later forms of the stories the scene is liable to be changed to ^ Murray, Rise of the Greek Epic, p. 197 (from Bethe, N.Jahrb., vii 672). 2/0 MYTH IN THE HOMERIC POEMS [CHAP. entirely different countries. Thus the fight of He?5inn and Hogni is located in the Orkneys in the Norse version of their story, while in Kudrun Hagen is made a king in Ireland. Again, in the Vitae Duorum Offarum the whole story of Offa and his single combat is transplanted to the English Mercia, the home of the hero's descendants. Yet in the case of Istros' story the requisite evidence seems to be altogether wanting. If the story really comes from local tradition it may very well be due to an imperfect acquaintance with the Homeric poems. But the context, which mentions Hector as well as Paris, suggests rather that Istros was referring not to the story of the Iliad at all but to an early adventure of the two brothers^ — presum- ably one of those accretions to the old heroic cycles, for which so many parallels can be found in late Teutonic authorities like Thit5reks Saga af Bern. The argument relating to the Locrian cult of Alexandra (Cassandra) need scarcely be considered at length ; for, however ancient this cult may have been, it was always connected with the sanctuary of Athene at Troy-. The cult of Hector at Thebes likewise seems to have been derived from the same quarter, perhaps in comparatively late timesl Again, the identification of Andromache's home with Thebes in Phthiotis is admittedly nothing more than a conjecture. One argument still remains for consideration, namely that the persons brought into contact with Hector come chiefly from the north-eastern parts of Greece and those encountered by Paris chiefly from Thessaly. Now it is to be observed that several of the persons whose names figure in Prof Bethe's lists {pp. cit., p. 670 ff.) are not said to come from Thessaly, Boeotia, etc. in the Iliad itself That they were derived from this quarter is merely an inference from the fact that other persons belonging to Thessaly, Boeotia, etc. bear the same names. It cannot for a moment be suspected that in v. 705, XV. 547 ff. the poets themselves were thinking of Orestes the son of Agamemnon or of Melanippos the famous Theban hero. 1 Cf. Crusius, S.-B. d. k. bayer. Akad., 1905, p. 774 f. 2 Cf. Bruckner, Troja und Ilion, p. 557 ff. 3 Cf. Crusius, op. cit., p. 761 ff. XIII] MYTH IN THE HOMERIC POEMS 2/1 But surely nothing can be more absurd than the proposition that persons who bear the same name must necessarily be identical in origin. In the Teutonic Heroic Age we know from historical sources of five kings named Theodric, all of whom were living within half a century of one another. Have we any reason for supposing that the ancient Greeks were more careful to avoid the use of names which had already been appropriated?^ Again, the lists given by Prof. Bethe contain merely a selection of the warriors encountered by Paris and Hector. The former fights in the Iliad not only with the Thessalian heroes Machaon, Eurypylos and Menesthios, but also with Menelaos from Sparta, Diomedes from Argos and Euchenor from Corinth. Hector's antagonists include, among others, Aias from Salamis, Stichios from Athens, Periphetes from Mycenae, Lycophron from Cythera and Amphimachos from Elis. On the other hand Melanippos is merely one, and by no means the most conspicuous, of the same hero's supporters. Considering the evidence as a whole therefore I fail to see that this argument is worth any more than the others. It appears then that the evidence adduced in favour of the theory which we are discussing is open to serious objection at every point^. But, leaving questions of detail, we have yet to notice that the theory as a whole consists of two main proposi- tions. The first of these is that the warriors of the Iliad are really ' tribal heroes,' i.e. in some sense or other personifications of tribes. In other words the contests described in the poem were originally conflicts of tribes and not of individuals. The second proposition is that these conflicts must have taken place between neighbouring tribes. It will be seen that this proposition ^ As a matter of fact the name Melanippos is borne by three Trojans and one Achaean in the Iliad. Nothing is stated regarding the home of the Achaean. The name Orestes is borne by one Trojan and one Achaean, besides the son of Agamemnon. * I have dealt with only one of the groups of names treated by Prof. Bethe ; but it is the one which he has discussed most fully. A second (Laconian) group is treated by him on p. 672 f. On this it will be sufficient here to refer to Crusius, op. cit., p. 771 fif., where it is pointed out that the chief argument rests apparently on a mistranslation. 272 MYTH IN THE HOMERIC POEMS [CHAP. depends very largely upon the first. If the conflicts of the Iliad really took place, as the poem states, between individual warriors or bands of soldiers, there is no occasion for supposing that the combatants were necessarily neighbours. As far back as the twelfth and thirteenth centuries Egyptian monuments testify, as we have seen, to enterprises far more distant than the expedition against Troy. Now the first proposition is of course nothing new in itself. On the contrary. Prof Bethe's theory is an outgrowth from a view which has been long and widely current — that the conflicts recorded in the Iliad are a reflection of the Aeolic settlement of the Asiatic coast. His purpose has been to show that only a comparatively small portion of the story comes from this source^, and that the bulk is derived from reminiscences of earlier tribal struggles in Greece, which have become embedded in the story of Troy. The two theories differ very greatly in the explanations which they give of the origin of various incidents and characters. But it is not to be overlooked that the principle of interpretation is the same in both cases. No one will deny that the personification of tribes and nationalities is to be found in both the poetry and the prose literature of many peoples. In certain passages of the Old Testament this principle of interpretation has been recognised from ancient times. But the authorities in which these passages occur cannot be described as heroic poems. Again Greek literature itself also yields plenty of obvious examples, such as the stories of Hellen and his sons and Danaos, several of which can be traced back to quite early poems. But these poems appear to have been of the Hesiodic, and not of the Homeric school. If we turn to the Teutonic peoples, evidence for such personification is abundant, and some of it belongs to our very earliest records. But, except in genealogical references such as we have dealt with above (p. 267), examples are not to be found in heroic poetry. The idea that the characters who are brought before us in the poems — let us say Beowulf or SigurSr or Witege — are themselves personifications of tribes is one which probably no scholar would entertain. 1 Cf. Neuejahrb., vn 662—9; xin 2 ff. XIII] MYTH IN THE HOMERIC POEMS 273 Now in the Homeric poems, as we have them, just as much as in Teutonic heroic poetry, the interest of the poets lies in the fortunes of individual heroes, not in those of the communities to which they belong. Even in those Teutonic stories which have the least claim to be regarded as historical there is no reason for doubting that such was the case from the very beginning. On the other hand the current hypothesis with regard to the origin of the Greek heroic stories postulates what can only be described as a complete revolution in the interests of the poets and their audiences. This however is a postulate which ought not to be accepted, unless decisive evidence is forthcoming in its favour. In the first place it must be observed that the existence of a poem or story which deals with reminiscences of tribal conflicts necessarily presupposes an absorbing interest in tribal history. It will probably be admitted by everyone that this interest can hardly have been of an academic character ; indeed, we may assume, I think, that it must be inspired by patriotic motives. If so, the foremost place will naturally be taken by that tribe or community with which the story originated. Now it is generally agreed that the Homeric poems contain both Aeolic and Ionic elements. Further, though opinions differ widely as to the relative importance of the two, there is a practical unanimity in believing that the Aeolic element is the earlier one. The tribal interest then, at least in the earlier elements in the poems, should be essentially Aeolic. For ' Aeolic ' we may practically say Thessalian (using the term of course in a geographical sense); for the Aeolic settlements, mainly at least, had proceeded from Thessaly. But Achilles, the chief hero of the Iliad, himself belongs to Thessaly ; and so all is well. Further, many scholars hold that the later or Ionic elements in the poem are marked by the introduction of Nestor. There was, apparently, a tradition current in Colophon that the inhabitants of that city had originally come from Pylos. Nestor therefore may be regarded as typifying the later or Ionic interests of the Iliad, just as Achilles typifies its earlier or Aeolic interests. This opinion however is by no means so widely entertained as the other. So much for the Iliad ; now let us turn to the Odyssey. Here we are confronted with a serious difficulty. Odysseus is 274 MYTH IN THE HOMERIC POEMS [CHAP. at least as much the hero of the Odyssey as Achilles is of the Iliad. But Odysseus belongs to the Ionian Isles ; and there is no trace of either an Aeolic or an Ionic population in these islands. It is not surprising then that the Odyssey is put aside by the advocates of the theory which we are discussing. The tendency is to regard it as a later work — originating perhaps at a time when tribal interests had become forgotten. We must confine our attention therefore to the Iliad. As applied to the Iliad the theory was long ago seen to be open to one serious objection. Achilles is the only one of the chief Achaean leaders who can be referred to Thessaly. His nearest neighbour is the Locrian Aias ; but the Locrians, in spite of their connection with Troy, cannot be regarded as an Aeolic people. All the other Achaean leaders who may be termed ' heroes of the first rank ' belong to the southern and western parts of Greece. Their positions geographically cannot be reconciled with the theory of Aeolic tribal wars. In order to obviate this difficulty various suggestions have been put forward. On the one hand we have Prof Bethe's hypothesis which brings the Trojan leaders, Hector and Paris, to the north-eastern parts of Greece. Upon this enough has been said above. On the other hand there is an older and still very popular hypothesis, according to which some of the southern leaders, Agamemnon in particular, originally belonged to the northern parts of the country. It is to this that we must now turn our attention. In the Iliad Mycenae is represented as being the home of Agamemnon. But it has been observed that this place is comparatively seldom mentioned, and that sometimes Agamemnon is said to rule over ' Argos.' Unfortunately there is a considerable amount of ambiguity in the use of the latter name. Occasionally it denotes the well-known city in Argolis ; but more often it is clearly used in a much wider sense, for the Peloponnesos or the whole of Greece. Once however (ll 68 1) we find the expression to HeXaajcKov "Apyo'i as a name for the home of Achilles. In ancient times the meaning of this expression was not known. Some authorities believed it to be the name of a city, while others understood it as a designation for the plain of Thessaly. Many modern scholars have adopted XIII] MYTH IN THE HOMERIC POEMS 2/5 the latter interpretation. But further, they hold that this was the original Argos, and that the application of that name to the Peloponnesos or any part of it is due to a misunderstanding on the part of later (Ionic) poets, by whose time the northern Argos had been forgotten. In favour of this view Prof Cauer^ brings forward the following arguments. " If Agamemnon, as well as Achilles, belongs to the oldest elements in the story, he also must come from a land in which Aeolic was spoken ; and indeed not Aeolic in the extended sense which the ancients gave the term — where the name includes Elean and Arcadian — but Lesbian-Aeolic. This was, as inscriptions show, the language of the original inhabitants of Thessaly." Again " Agamemnon started with his fleet from Aulis....He was associated with Achilles in the story from the beginning. His Argeioi are the companions of the Achaeans led by Achilles. The two tribal names are used for one another indifferently, and either of them can be employed as a designation for the forces which fought at Troy. Conse- quently the Argeioi and the Achaeans must have been neighbours." Further, it is urged by Prof. Cauer that the epithet iTnro^orov as applied to the Peloponnesian Argos is inappropriate. This state possessed no cavalry in the fifth century, and none of consequence at any time. On the other hand it is a very suitable epithet for the plain of the Peneios ; Thessalian cavalry were famous. Lastly, it is argued that the expression Kad' (dv) 'KWd8a koI jxeaov "Apyojpr]s. ^ This passage seems to indicate that the poet included the basin of the Spercheios in Phthia. 280 MYTH IN THE HOMERIC POEMS [CHAP. the Catalogue which can be identified with certainty, namely Trechis. On the whole also they favour the view that ' Alope ' is the Phthiotic Alope. Achilles' country then is the basin of the Spercheios, together probably with the coast lands on the Malian Gulf between the mountains of Oite and Othrys. It may have included the northern slopes of the latter, in the territory of the Dolopes ; but neither the evidence of the Catalogue itself nor references in other parts of the Iliad give us any warrant for supposing that it extended into the plain of Larissa\ Now there is not a particle of evidence that an Aeolic dialect was ever spoken either in the basin of the Spercheios or in the districts bordering on the coasts of the Malian Gulf. A form (Mcr^^aeio?) which may be an Aeolic patronymic occurs in an inscription from Melitaia- ; but the inscription in other respects is definitely non- Aeolic, although Melitaia lies to the north of Othrys. A similar form (Ku^ioTeta) occurs in an inscription found near Pteleon^ which is too short for us to determine its character ; but Pteleon lies in the extreme east of Phthiotis, outside the Malian Gulf, and according to the Catalogue belonged to Protesilaos. The only inscription of definitely Aeolic (Thes- salian) character as yet found in Phthiotis comes from near Eretria, just inside the boundary. This place is not mentioned in the Catalogue, or elsewhere in the Iliad ; but Strabo (ix 5. 10) conjectured that it belonged to Achilles' territories. If we are to follow the indications given by the Catalogue the question would seem to lie between the territories of Protesilaos and ^ So far as I can see, the identification of ' the Pelasgian Argos ' with the plain of Larissa rests merely on a conjecture, of the truth of which even Strabo himself {IX 5. 5) was not confident. It may very well have been suggested by the name (Pelasgiotis) borne by this district in later times. The oracle quoted by Prof. Meyer (Forsck., p. 30, note i) proves nothing, and there is no evidence that the name was used in historical times. On the other hand the fact that the citadel of the Pelo- ponnesian Argos was called Adptcra (cf. Pausanias n 24. i) does suggest a connection between this name and 'Apyos. But if so, it is more natural to think of Larissa Cremaste, which according to Strabo (ix 5. £3, 19) was also called Pelasgia. This place is much nearer to the Spercheios than the northern Larissa, and in spite of Strabo (ix 5. 14) may quite possibly have been included within the same territory. 2 Cauer, Delectus^, No. 388. * ib.. No. 390. I XIIl] MYTH IN THE HOMERIC POEMS 28 1 those of Eurypylos ; there is nothing to show that Achilles' country was believed to extend so far. It is most surprising therefore to find Prof. Cauer^ concluding from such evidence as this that " we are justified in claiming the valley of the Spercheios also as an Aeolic district and Achilles as a hero of Aeolic nationality." For the language of Achilles' country itself we are by no means without evidence. Fairly long inscriptions have been found at Hypate in the valley of the Spercheios and at Lamia, to the north of the Malian Gulf — to which may be added an inscription, apparently of the Oitaioi, at Drymaia in Phocis. All these show the form of language usually known as ' north- west Greek,' and the same is true of other inscriptions found in the north and east of Phthiotis. Although they are all late, there is no valid reason for doubting that this language is indigenous^ Only two inscriptions^ so far as I know, contain references to Aetolian magistrates. The dialect is almost identical not only with that of the Aetolian inscriptions, but also with those of the Locrians and Phocians, the former of which is well known from much earlier times. Greek communities as a general rule were slow to change their language, and the influence of the Aetolian League was scarcely of such a character as to favour the permanent extension of its dialecf. ^ Grundfragen der Homerkritik^, p. 214. 2 By this of course I do not mean that the inscriptions give an absolutely faithful reproduction of the local pronunciation, any more than do those of the Aetolians. From the fourth century onwards ' phonetic spelling' appears to have been superseded in most parts of Greece. No earlier inscriptions, representing the Achaean dialect in its purity, have been found as yet. From Thetonion however, near Cierion in Thessaliotis, about twenty miles north of the border, we have an inscription of the fifth century (C. /. C, xii ii 257) in a curious mixed dialect, which combines north- western Greek and Thessalian (Aeolic) characteristics in the proportion of about 7 : 3 (or 4). The evidence of this inscription seems to me to dispose definitely of the hypothesis that the introduction of north-western Greek into this region was due to the influence of the Aetolian League. Even the dialect of Pharsalos is not quite pure Thessalian. 3 Cauer, Delectus'^, Nos. 239, 386. The former is included by Prof. Cauer, no doubt rightly, among the Aetolian inscriptions. * The extension of the Dat. pi. ending -ots to consonant-stems etc. has often been quoted as a mark of Aetolian influence. But in reality it is common to all the dialects of western Greece. The earliest examples apparently occur in Elean and Locrian inscriptions. f 282 MYTH IN THE HOMERIC POEMS [CHAP. If we are to trust all the evidence which we possess Othrys, and not Oite, was the southern limit of the Aeolic (Thessalian) dialect. The communities of Phthiotis were politically dependent upon Thessaly, but they seem never to have been subjugated in the same way as the Aeolic population north of the mountains^ They had their own troops and sent a separate contingent to Xerxes' army. Indeed Herodotus (vii 173, 196 ff.) clearly distinguishes between 'Thessaly' and 'Achaia' (i.e. Phthiotis). Moreover it appears to have been the general opinion among the Greeks themselves that this district was the original home of the Achaeans of the Peloponnesos. Pausanias (vii 1.6) traces the Peloponnesian Achaeans to Archandros and Architeles, the sons of Achaios, who had come from Phthiotisl Herodotus (11 98) was evidently familiar with some form of this story, though he calls Archandros son of Phthios and grandson of Achaios. The supposed connection therefore goes back at least to the fifth century. How far these genealogies were constructed upon linguistic affinities is a question which needs some dis- cussion. We may remark in passing however that the dialect of the Peloponnesian Achaia, so far as it is known to us*, shows but little difference from the dialects north of the Gulf of Corinth. It is commonly included in the list of ' north-west Greek' dialects. In the meantime we may notice an argument which has ^ According to the generally accepted view, which there seems to be no reason for doubting, the language of the Thessalian (Aeolic or ' North-Thessalian ') in- scriptions belonged originally to the indigenous population. The name ' Thessalian ' however, properly speaking, belonged to the invaders, regarding whose language we have no information. '^ Strabo (viii 5. 5) likewise connects the Peloponnesian Achaeans with Phthiotis; but he attributes their settlement in the Peloponnesos to an invasion by Pelops. In this passage (as in many others) it is greatly to be questioned whether Strabo (or the authority whom he followed) was recording genuine tradition — whether he was not rather endeavouring to provide an explanation of the traditions. •* The inscriptions are late ; but their evidence as to the general character of the dialect is confirmed by some short but early inscriptions from the Achaean settlements in Italy. It is assumed by many scholars that Arcadian was the original language of the Peloponnesian Achaeans ; but I am not aware that any evidence worth con- sideration has been adduced in support of this view. No ancient authorities, so far as I know, connect the Arcadians with the Achaeans, nor do the Arcadians themselves appear to have claimed such a connection. XIII] MYTH IN THE HOMERIC POEMS 283 sometimes been brought forward in support of the hypothesis that Aeolic was once spoken much further south. This is the presence of an Aeolic or semi-Aeolic form of language in Boeotia. The ancients themselves believed that the Boeotians were not indigenous. Thucydides (i 12) states that they had been expelled by the Thessalians from ' Arne ' after the Trojan War ; but this was not the only form of the story. On the other hand many modern scholars have adopted the view that the 'north-western' dialects of Locris and Phocis were intrusive\ and that the non-Aeolic characteristics of the Boeotian dialect itself were due to an extension of the same movement — in short that Aeolic was the earliest form of Greek spoken throughout the whole region from Thessaly down to the borders of Attica. For such a displacement of population^ no evidence is to be found either in history or tradition. Moreover this hypothesis has opposed to it the evidence of what may be called linguistic geography. The Ionic dialects of Euboea and Attica have much more in common with the ' north-western ' dialects of Locris and Bhocis than they have with Boeotian. The latter indeed stands quite isolated in many respects among the dialects of this part of Greece. It will be sufficient here to notice the close pronunciation of e (e.g. in Gen. sg. fieiv6v 'A6avo8copiov Ava^rjcrivedis re Aficf>iaX6s B' vlos TloXwrjov TfKTOviSao • av 8e Koi "EvpvaXos jipoTokoiyco i(Tos Aprji, 'Sav0o\i8r]s 6\ K.T.X. Other Phaeacians have names of the same type, e.g. Nauo-Z^oo?, ^avaiKaa, "AXio^, KXfToi^r^o?, 'E^ei'T^o?, YIovtovoo<;. In such a case as this the poet can scarcely have intended to deceive his audience. Indeed the principle is clearly admitted in another passage (XXIV 305), where Odysseus in a false story describes himself as vi6<; 'A^etSayro? UoXvirrj/xovlSao avaKTo, is believed to be identical with a certain prince named MTroybavos (lloydavos), who, according to Laonicos (p. 28), was granted by Dusan the territories between Pherrai and the Axios (Vardar), and who about 1372 submitted to Murad together with the other Servian princes in this region. To these we may add the vojvoda Vladeta ; for there can be little doubt that this person is to be identified with that Vlathico Vlagenichio who, according to the anonymous translation of Ducas' history (p. 352), was sent by his uncle luathco (Tvrtko), king of Bosnia, to support Lazar with 20,000 men. On the other hand some doubt has been e.xpressed with regard to Milos Obilic (or Kobilovic) the chief Servian hero of the story. He is not men- tioned apparently by any strictly contemporary authority. Yet the traditional account of Murad's death is known to the two Greek historians Ducas and Laonicos, both of whom are believed to have written within about three quarters of a century after the battle. The latter (p. 54) states that accord- ing to the Greek version of the story a Servian nobleman named Mt)Xois rode fully armed into the Turkish camp, representing himself to be a deserter. Murad gave orders that he was to be allowed to come near and say what he wished. But when he reached the door of the Sultan's tent he threw his spear and slew Murad, meeting with his own death immediately afterwards. Laonicos however also says that the Turks gave quite a different account of the affair — namely that as Murad was pursuing the enemy a Servian {av8pa Tpi/3aXX6i/), who was on foot, turned and transfixed him with a javelin. Ducas' version of the story (p. 15) resembles that given by the poems in the fact that the assassin uses a dagger. His name is not given, but he is said to have been a young and distinguished Servian and to have asked to see Murad as a deserter with important information. Closer aflfinity with the poems is shown by the anonymous translation of Ducas' history (p. 352 ff.), which contains much additional matter. Indeed it is scarcely possible to doubt that the additions are partly derived from poems, though these may not have been exactly identical with any which are now extant. When Milos ('Milos Cobilichio, capetanio de Lazaro') reaches the Sultan's tent we are told, as in the poems 2, that he is bidden to ^ I do not know whether this is historically correct. There seem to be a number of historical references to Milica, Lazar's queen; cf. Engel, op. cit., pp. 311, .=531. .S46f. * Cf. Pasic, op. cit., vi v. 13 fif. (p. 92). This passage is taken from a Croatian poem (Nr 6, v. 166 ff.) published by Miklosich, Denkschriften d. k. Akademie d. Wissenscha/ten (Vienna), xix p. 73 fif., from a MS. collection at Ragusa dating from about 1728. Mme Mijatovich's poem on the same subject (p. 120 ff.) differs a good deal from this and bears a closer resemblance to the Italian in one or two points. XIV] THE BATTLE OF KOSSOVO 315 kiss Murad's foot 'according to the usage of his kingdom'.' Far more striking however is the account of the banquet on the preceding day, which in places appears to be little more than a free translation of a fragmentary poem published in Karadzic's collection 2; "El zorno precedente a quelle che segui la iniqua et infelice bataglia, Lazaro convocati tutti i signori et principali del suo imperio^, comando che se aparechiasse una sdraviza secondo la usanza dela sua corte ; in laquale, come gratioso et benigno signore, a tutti porse la sdraviza con sua mano. Quando la volta tocco a Milos, se fh dar una grande taza d' oro piena de pretioso vino* ; la qual porzendoli disse a Milos : ' Excellentissimo cavalier, prendi questa sdraviza, che con la taza te dono...sdravize per amor mio. Ma molto mi doglio che ho inteso una mala novella, che al tuo dispoto sei facto ribello'^ Al qual Milos, reverentemente presa la taza con chiara faza, disse : ' Signor dispoto, molto te ringratio della sdraviza et taza d' oro che m' ai donata. Ma molto mi doglio dela mia dubitata fede^. Doman de matina, se dio dark efifecto al' alto pensier mio, se cognoscera se io son fidele o ribello dela tua Signoria." This translation is believed to be of Dalmatian origin'' and, according to Prof. Bury®, itself dates from the fifteenth century. We are bound to con- clude therefore that Milos' exploit was treated in poems — from which some of the extant pieces are ultimately descended — within a century of the battle. The earliest direct reference to poems dealing with Khobilouitz (i.e. Milos ' 'Secondo la usanza del suo imperio' (cf. Miklosich, 6. 167 : ' Ovaki su zakoni, Milosu, u zemlji mojoj,' etc.). ^ Vuk Stef. Karadzid, Srpski Narodtie PJesine, Vol. il (Vienna, 1875), 50 iii (p. 310 ff.). This is translated, with a few slight changes, in Mme Mijatovich's piece 'The Banquet before Battle' (p. 116 ff.). A somewhat different account of the same incident — and showing less resemblance to the Italian — occurs in Miklosich, 6. 116 ff., a passage which is not used by Pasic or Mijatovich. ^ Cf. Karadzic, li 50 iii, v. 3 f. : Svu gospodu za sofru sjedao {scil. Lazare), svu gospodu i gospodidide. * Ibid., V. 13 : 5 Ibid., V. 31 ff. ® Ibid., V. 39 ff. Car uzima zlatan pehar vina, etc. Zdrav Milosu, vjero i nevjero ! prva vjero, potonja nevjero ! Sjutra ces me izdat' na Kosovu, i odbjeci Turskom car-Muratu ; zdrav mi budi ! i zdravicu popij : vino popij, a na cast ti pehar. Vala tebe, slavni knez- Lazare ! Vala tebe na tvojoj zdravici, na zdravici i na daru tvome ; al' ne vala na takoj besjedi ; etc. ' Cf. Pasic, op. cit., p. 26. * Gibbon's Decline and Fall of the Roman Empire, vii p. 327. 3l6 THE BATTLE OF KOSSOVO [CHAP. Kobilovic) goes back to the first half of the sixteenth century, at which time they are said to have been numerous ^ It may be however that the accounts given by Ducas himself and by Laonicos are also derived ultimately from poems 2. But if so the poetic treatment of the subject must have begun within living memory of the battle ; for it is clear from the evidence of these authors that by the middle of the fifteenth century the story had come to the Greeks in more than one form and that it had even attained great celebrity among them^. The Kossovo poems certainly give us some evidence for the phenomenon which we have above (p. 307) called 'transference.' In one poem (Karadzic, II 46, v. 59 ff.) we find it stated that King Vukasin entered into the battle and met with his death. In reality he was killed at (or shortly after) the battle of the Marica in 1371. The same poem also, immediately afterwards (v. 71 ff.), introduces Erceg Stepan, who lived nearly a century later. But this poem* is of a very peculiar type — distinctly non-heroic — and it is not legitimate to draw conclusions from it as to the poems in general. It would be of far more importance for our purpose if we knew that Jug Bogdan was attracted into the story. He must certainly have been an old man in 1389, since he had his territories granted to him by Dusan, who died in 1356. But he is regularly described as 'the old' {start) in the poems. The fact also that he had been a vassal of Murad from 1372 onwards is hardly conclusive; for it is clear that a great effort was made in 1389 to unite the various Servian princes. On the whole the evidence, so far as it is known to me, does not seem to indicate that attraction or transference has played an important part in the story. For the invention of characters there is, so far as I know, no absolutely decisive evidence. But probably few would be inclined to regard such a person as Vaistina, the servant or squire {sluga) of Music Stefan^ (Karadzic, ' Cf. Murko, Geschichte der dlteren siidslaw. Litteraiitren, p. 205. - This is perhaps rather suggested by one or two of Laonicos' sentences, e.g. (MtjXoij' (paffiv) CoirkKTiiivov iXavveiv avv t<^ 'ivirifi iirl rb' Afiovpareoj aTpaTbireSov, clis hv avTo/jLoKovvTa dird tuv evavTiwv (p. 54) ; and more especially by the speech which Ducas (p. 15) attributes to Milos: "Boi^Xo/itat tovtov iSelv koX \6yovs rivas inro^idvplffai COS iyKpaTT]s yeviffdai rovTovt toO Tro\i/j.ov ^veKa yap tovtov avT6fJ.o\os i\ri\v6a." With the last sentence we may compare Miklosich, 6. 164 : Ja sam ti se odvrg'o od vojske Lazara kneza. ^ Cf. Laonicos, I.e.: "E\\Tifes...\4yov(nv avSpa yewaiOTaTov ideXrjffai inbvra viroaTTJvai dyiUva KoXKicrTov 8ri tCjv irujiroTe yeuofj-efup, and again, below : opfiTjv iracwv 8ri KaWi(TTT)v (if i]/jLeh ifffiev. * The first part of this poem (vv. i — 44) corresponds to Mijatovich, p. 104 ff. ; the second part (v. 47 ff.) to Mijatovich, p. 126 ff. In such cases as this it is much to be regretted that Mme Mijatovich did not adhere to Karadzic's text. If the explanation be that her version is derived from a different source — and not due to arbitrary transposition — this should have been made clear. ^ In Miklosich, 6. i — 58, which deals with the same story the squire is called Oliver, while his master's name appears as Busic Stjepane. I XI V] THE BATTLE OF KOSSOVO 317 II 47), otherwise than as a product of fiction ; and the same remark applies to the squire Milutin who brings the news of Lazar's death to the queen {lb. 45. 146 fif.). It is a more difficult matter to form an opinion as to whether any of the more important characters are fictitious. Several of them appear to be unknown from contemporary historical works. Among these we may mention Music Stefan himself and more especially Banovic Strahinja, the hero of the longest poem in the cycle lyib. 44). So far as I know, the earliest reference to the latter is in the Chronicle of Tronosa, which mentions inci- dentally the destruction of his palace 1. Again, according to the poems Milos was accompanied on his errand to Murad by two of his friends, Milan Toplica and Ivan Kosancic. Both these persons are mentioned in the same connection by the Chronicle of Tronosa, which adds that at the banquet on the preceding night they, as well as Milos, had been charged with disloyalty by the king. This chronicle is believed to be derived from a MS. of the sixteenth century; but there is practically no doubt that it has drawn largely from poetic sources. Our earliest authorities, Laonicos and Ducas, together with the translation of the latter, seem to imply distinctly that Milos carried out his exploit alone. That however does not prove that the characters themselves are fictitious. On the whole, considering the limited amount of information which early records furnish, it would probably be wise to hesitate before adopting the view that any of the more important characters are invented — at all events those which can be traced back to within two cen- turies of the battle. There seems to be no evidence for the introduction of what can properly be called mythical beings in poems of the Kossovo cycle ^. But sometimes we certainly find supernatural incidents. In Karadzic, li 45. 119 ff. two crows from the field of Kossovo bring to the queen the first news of the battle and of Lazar's death. A more extravagant case occurs in the opening verses of {ibid.) ll 46 — a poem to which we have already referred. Lazar is here made to receive a letter dropped by a swallow (which is carried by a falcon) offering him the choice between the heavenly and earthly kingdoms. In another (Croatian) poem (Karadzic, 11 48^) the mother of the Jugovici prays that she may receive the eyes of a falcon and the wings of a swan. Her prayer is answered, and she flies to the field of Kossovo and sees the dead bodies of her sons and husband. This poem is largely taken up with the marvellous throughout and has little in common with heroic poetry. But it is from the point of view of their presentation of the story that the Kossovo poems chiefly merit our attention. This presentation contains many features which may be included under the head of fiction. Yet in ^ Quoted by Pasic, op. ctt., p. 30. 2 Other poems, both Christian and Mohammedan, frequently introduce Vile, i.e. nymphs or elf- women. ^ Not included in the collections of Pasic and Mijatovich. A slightly variant form (apparently of Montenegrin origin) is published, together with a translation, in Krauss' Slavische Volkforschungen, p. 289 f. 3l8 THE BATTLE OF KOSSOVO [CHAP. certain cases it is a question whether we have not rather what may be called a growth of myth. By this I mean the introduction and development of motives which, though incorrect historically, can hardly be regarded, at least in their entirety, as conscious inventions of an individual. They would seem rather to have originated in rumour and popular misconceptions. As examples we may take what are perhaps the two most salient features in the story — the exploit of Milos and his confederates and the treachery of Vuk Brankovic. We have already dealt with the first of these incidents. The poems make Milos and his companions perform prodigies of valour before they are overcome. But our earliest authorities state that Milos was killed almost immediately; and they imply that he was alone. Indeed we know that the Turks gave quite a different account of Murad's death. It is not at all clear that between the two the Servian account possesses the greater probability. But it may very well have been believed among the Servians from the very beginning, whether its origin is to be traced to genuine information derived from the Turkish camp or merely to idle rumour. Again, the treachery of Vuk Brankovic is proclaimed again and again in the poems. But there is no evidence earlier than the sixteenth century to substantiate the charge^. The first reference to treachery in Lazar's army occurs in the translation of Ducas' history (p. 354) ; but here the traitor is called ' Dragossavo Probiscio, capitaneo del campo del dispoto.' It has been suggested that Vuk's unenviable celebrity in the poems is due to the unpopularity of his son, George Brankovic, who ruled over Servia from 1427 to 1457^. The above brief sketch will probably be sufficient to show that these poems are capable of throwing a good deal of light upon the origin and development of a heroic story. The period is one for which, comparatively speaking, a fair amount of information is available ; and quite possibly more might be obtained by a careful investigation of the documents of that age. It is to be remembered of course that after the middle of the fifteenth century nothing like court poetry can have existed in Ser\na. We should expect then that from this time onwards the poems would become more and more permeated by those characteristics which we have assigned to Stage in of our scheme. As a matter of fact some of the poems show these charac- teristics only to a comparatively slight degree. This is especially the case with the poem on the banquet (Karadzic, 11 50 iii) ; but part of this poem, as we have seen, can be traced back to the fifteenth century. On the other 1 Prof. Murko (op. ciL, p. 202) states that Vuk "in der Schlacht in hervorragender Weise seine Pflicht erfiillt und sich dann mit den Turken gar nicht ausgesohnt hat wie Lazar's Sohn Stefan." I do not know the authority for the first part of this statement ; but according to the Turkish account (cf. Engel, Geschichte von Serwien, p. 346) the (right) wing of the Servian army which was commanded by Vuk was successfiil. ^ Cf. Pasic, op. cit., p. 42. XIV] THE BATTLE OF KOSSOVO 319 hand the characteristics of Stage iii are very strongly marked in Karadzic II 46 and 48. I should expect that these poems are late compositions. So far as I have been able to deal with the subject the result of the discussion has been that there is little or no definite evidence for the inven- tion of characters. That is a result which can scarcely be regarded as surprising, in view of the history of Servian poetry. On the other hand the conditions were such as we should expect to be exceptionally favourable to the development of transference or attraction. Yet there is but little satis- factory evidence in this direction. I am inclined therefore to think that the force of this principle has been considerably overestimated by recent writers. What the Kossovo poems do seem to suggest is — not that the characters of the Iliad were invented or attracted from other quarters, but that their exploits and their relations with one another may in reality have been very different from what we find depicted in the poem. CHAPTER XV. THE COMMON CHARACTERISTICS OF TEUTONIC AND GREEK HEROIC POETRY. •^ Our review of the Homeric poems has led us to conclude that their origin and early history was in many respects analo- gous to that of the English heroic poems ; and further, that there is no valid reason for regarding the stories with which they deal as mythical or fictitious, although we cannot, as in the case of the English poems, actually prove that they rest upon a historical basis. We must now endeavour to see what common elements the two series of poems contain in regard to style and spirit. This will enable us to determine whether the term ' Heroic Age,' as applied to the two cases in common, can be held to mean anything more than an age of ' heroes,' whose deeds were celebrated in poetry. The most cursory glance at the two groups of poems will be sufficient to show that they contain many common features in regard to style. In both we find the constant repetition of the ^ame formulae, e.g. in the introduction of speeches. Thus in the first part of Beowulf eight speeches out of thirteen by the hero himself are introduced by the formula : Beowulf map elode beam Ecgpeowes, while three of Hrothgar's seven speeches follow the words : Hro^gar inajyelode helm Scyldinga. In the Iliad we may compare the constant repetition of such formulae as : rov 8' a'Tra[JbeL^6fievoegn, i.e. conies ; ava^ — dryhten, \.q. princeps), seem to correspond to one another almost exactly, though it is not easy to find a satisfactory render- ing for them in modern English. For kings too, especially such as are of preeminent position, we find in each case a very similar expression (ava^ dvBpwv — eor/a dryhten\ which properly denotes the relationship of liege- lord. In both cases alike the leading idea of the Heroic Age may be fittingly summed up in the phrase /cXea dvhpwv. This is practically equivalent to the Anglo-Saxon doni, with which we have dealt above (p. 87 f.). It is essential to notice that the object so much prized is personal glory. In Hector's speech before his combat with Aias (cf. p. 325) the glory which would result from the combat to the hero or his opponent is the only subject touched upon. No consideration is taken of any effect which might be produced thereby upon the fortunes of the war. The same characteristic appears throughout the passages which we have quoted and countless others. Occasionally we hear also of pride of family, but scarcely ever of any truly national feeling. Patroclos exhorts his men to bravery (II. xvi 270 ff.) in order that they may win glory not for the Achaean nation but for their own personal lord ; and he adds further that by so doing they will bring shame upon the national leader. Achilles himself retires from the conflict owing to a personal wrong, and only returns to it in order to avenge his friend. The same phenomena appear in the English poems. That Wiglaf whose bravery is said to be ' inbred ' (cf. p. 328) was the son of a certain Weohstan, whose great achievement was the slaying of the Swedish prince Eanmund. Yet Eanmund was at this time apparently under the protection of Heardred, king of the Geatas, ^ iam uero infante in omnem uitam ac probrosum stiperstitem principi suo ex acie recessisse. ilium defendere, tueri, sua quoque foriia facia gloriae eius assignare prae- cipuum sacramentum est. principes pro uictoria pugtuint, comites pro principe. 330 THE CHARACTERISTICS OF HEROIC POETRY [CHAP. who also lost his life in the same war. Weohstan however, though he belonged to the Geatas, was in the service of Onela, their enemy. It would seem then that he was fighting against his own nation. Such cases appear to have been by no means uncommon in the Teutonic Heroic Age. For it was customary at that time for young noblemen to take service under foreign princes; and the obligations which personal service imposed were held to be superior to all others. Love of home and zeal in its defence are of course frequently mentioned in both groups of poems. We may refer to the common phrase (fyiXrjv eV warpiha a\a.ptj}v Kdfffios Ik re rod ttLXov /cat rod ddparos aXovpyds re kcu aWws /SaeriXet irpiiruv iKpi/xaro Oavfiaaros offos. 342 THE CHARACTERISTICS OF HEROIC POETRY [CHAP. not only after a general victory but also in incidental and more or less private forays \ it will be seen that warfare of the heroic type offered very substantial inducements, apart from the acquisition of glory. It is not to be supposed that such warfare was really of an effective character. Even in the Iliad itself (VI 67 ff.) a warning is raised against it by the old Nestor, who is represented {ib. II 555) as an exceptionally skilful general, but his advice seems to be unheeded. Between two armies of the heroic type the issue had to be decided, if at all, by a pitched battle. Sheltered behind fortifications, even an inferior army had not much to fear. In that case the people most exposed to danger were what we should call non-combatants — not only the women and children and unwarlike dependents of the combatants them- selves^ but also any neighbouring communities who might be caught unawares by bands of hungry warriors. On the other hand against organised forces, like those of the later Spartans or the Romans, at all events if commanded by generals who followed a definite plan of campaign, the Homeric Achaeans would have had no chance after the first encounter. This may be seen especially from the history of the campaigns against the Italian Gauls, who appear to have employed a very similar method of warfare. The Saxons and other northern peoples owed their successes largely to the rapidity of their movements, combined with the fact that they had command of the sea. But, so far as we know, they seldom or never had to deal with any considerable Roman army. In Britain their pro- cedure was probably similar to that followed by the Scandinavian invaders in the Viking Age ; but the latter were unable to cope ^ Even in time of peace merchants might turn into freebooters (cf. Od. xiv 262 flf). Piracy indeed was scarcely regarded as disreputable {ib. ni 72 ff. ; cf. p. 322). The same conditions prevailed during the Viking Age and doubtless also during the Teutonic Heroic Age. * As illustrations of the barbarities associated with warfare of this type we may refer to the speech of Theodric, king of the Franks, given by Gregory, in 7, and to the behaviour of Theodberht's army in Italy, recorded by Procopius (Goth. Ii 25). At such a time the atrocities which Greek tradition relates in connection with the fall of Troy would have caused little comment. Yet the early heroic poems give less evidence even than the Homeric poems for cruelty of this kind. XV] THE CHARACTERISTICS OF HEROIC POETRY 343 with prolonged and organised resistance, such as was offered by Alfred the Great. The armies of the Goths at the height of their power were doubtless more formidable ; but they had probably learnt much from long experience of the Romans, both as foes and allies. It is not to be forgotten however that supremacy came ultimately to the Franks, who of all the Teutonic peoples seem to have been least affected by the spirit of the Heroic Age. ' CHAPTER XVI. SOCIETY IN THE HEROIC AGE. The evidence of the German poems for the social and political conditions of the Heroic Age cannot be regarded as trustworthy owing to the lateness of the period in which they were composed. In principle the same is true also of the Norse poems. These reflect the conditions of the Viking Age rather than those of the Heroic Age, though, as we have already noted, the difference here is less marked. On the other hand, in addi- tion to the Anglo-Saxon poems and the works of contemporary Roman historians, such as Ammianus Marcellinus, Jordanes and Procopius, we have valuable evidence from the early Teutonic codes of law. Some of these, such as the Lex Salica and the Lex Burgundionum, date from the first half of the sixth century or earlier, i.e. from the Heroic Age itself, while a number of others — in particular we may note the earliest English laws — belong to the following two centuries and show probably little deviation from the custom of the Heroic Age. All the codes of course contain certain Roman or Christian elements ; but this influence in some cases goes back to the fifth century or even further. The chief forces which governed the social system of that age were the bonds of kinship and allegiance. The influence of the former extended not merely, as with us, to rights of succes- sion and duties of guardianship over children and women. It was also the power by which the security of the property and person of each member of the community was guaranteed. If a man received injury or insult, his kindred were bound to assist him to obtain redress. If he were slain they had to exact CHAP. XVI] SOCIETY IN THE HEROIC AGE 345 vengeance or compensation from the slayer. On the other hand not only the slayer himself but every member of his kindred became liable to vengeance, and each had to pay his quota towards the compensation (wergeld), just as it was divided among the kindred of the slain — the proportion varying in both cases according to the degree of relationship. In case of blood- shed a certain sum had also to be paid to the king, even in the earliest times of which we have record ; but this sum seldom exceeded half the wergeld, and as a rule amounted to con- siderably less. The character and size of the kindred appear to have varied in different nations. Some laws speak of claims to succession as remote as the seventh degree, while the rights and duties connected with the payment and receipt of wergelds seem gene- rally to have extended as far as third cousins, i.e. the descendants of great-great-grandparents. Again, we hear sometimes of royal or noble families which bore a common name derived from some ancestor, real or mythical, from whom their power or preroga- tives were believed to be inherited. Such were the Oescingas, the Wuffingas and the Icelingas, the royal families of Kent, East Anglia and Mercia respectively ; so also the Scyldungas (Skioldungar) among the Danes, the Merovingi among the Franks, and the Agilolfinga and other noble families among the Bavarians. Persons belonging to these families had probably — in some cases certainly — special wergelds ; and the throne or principality seems to have been regarded as in some sense family property. Some writers believe that kindreds in general were permanent organisations of this kind, and that originally they held land, and possibly other property also, in common. But this view goes a good deal beyond what the facts warrant, at all events for the period with which we are dealing. It is clear that at this time kinship on both sides was recognised everywhere ; maternal relatives shared in the payment and receipt of wergelds with those on the father's side, though not always in the same proportion. Moreover the idea that the inclusion of the maternal relatives was due to an innovation cannot be maintained. Thus, in spite of the fact that the Prankish kings claimed the throne by direct descent in the male 346 SOCIETY IN THE HEROIC AGE [CHAP. line from Merovechus, there are clear indications that Prankish law in its earliest form gave priority to the mother's side. No doubt on the whole the agnatic system of relationship had become predominant almost everywhere in the Heroic Age ; but sufficient traces of the opposite system remain to render it probable that a change had taken place not so very far back^ Any such change of course involves — or rather perhaps implies — a weakening in the force of the bonds of kinship ; and of this we have very clear evidence in the Heroic Age. Now it has often been pointed out that early Teutonic custom seems to have made no provision for the case of homicide within the kindred. In such a case the persons on whom vengeance devolved would be identical, in part at least, with those against whom it would be directed — and so also with the compen.sation. It is generally held that homicide of this kind was extremely rare and that, when it did occur, the slayer was outlawed. That would no doubt be in accordance with primitive custom. Indeed in a state of society based on blood-relationship the life of a kinsman must be sacred above all else. Further, it is clear enough that the shedding of kindred blood was regarded with abhorrence in the Heroic Age. Thus Procopius {Goth. II 14), describing the euthanasia practised by the Heruli, states that when the dying man has been laid upon the top of the pyre, one of his countrymen goes up with a dagger and stabs him ; but he adds explicitly that this man must not be related to his victim". Again, to take another point of view, perhaps the saddest passage in Beowulf is that which relates how Herebald was accidentally killed by his brother Haethcyn. But the aspect of the case which first strikes the poet is not one which would appeal to a man of modern times. " That was a slaughter without compensation," he says (v. 2441 ff.), "the prince had to lose his life unavenged." Yet, in spite of all this, instances of the slaying of kinsmen seem to have been by no means uncommon in the Heroic Age. ^ Cf. The Origin of the English Nation, p. 327 ff. 2 ^vyyevi) yap avrij) rov (povia. elvai ov difiit. Cf. Greg. Tur., II 40, where Clovis says : nee enim possum sanguinem parentum meorum effundere, quod fieri nefas est. But this is represented as mere hypocrisy ; cf. 11 41, ad fin. XVl] SOCIETY IN THE HEROIC AGE 347 In Beowulf the spokesman of the Danish kings, Unferth, is said to have killed his brothers, and though the fact was a reproach to him, it apparently did not prevent him from holding an im- portant office at court. In the same poem we hear of dissensions within the Swedish royal family, which ended in death for both Onela and Eanmund. According to the legends preserved in Ynglingatal this family had had a very bad record for such quarrels in the past. Among the Goths we have the case of Eormenric, who put his nephews Embrica and Fritla to death. And it is by no means only in poetry or tradition that we meet with such cases ; historians also furnish numerous examples. Thus according to Gregory of Tours (ll 28) the Burgundian king Hilperic was killed by his brother Gundobad, while Sigis- mund, son of the latter, had his own son, Sigiric, put to death (ill 5). The Thuringian king Irminfrith slew his brother Berht- hari (III 4) ; the Prankish king Sigiberht was murdered by the orders of his son Hlothric (ll 40). Clovis is said to have put to death a number of his relatives, while his sons and grandsons were repeatedly involved in deadly strife ^ In view of such evidence we must conclude that the primitive sanctity of the family was giving way in the Heroic Age. For the change of feeling which was taking place one passage in Beowulf is particularly instructive. In the struggle between Onela and Eanmund the latter was slain by one of the king's knights named Weohstan. He stripped the dead man of his arms and brought them to Onela who presented them to him and "said nothing about that deed of guilt although it was his brother's son whom he (Weohstan) had laid low^" To the modern reader the poet's reflection seems strange ; for Onela had been relieved of a dangerous foe, who was trying to deprive him of the kingdom. Yet there can be no doubt that according ^ In some cases the deed was certainly done by the relative's own hand. Such was the case with Lothair and the sons of Chlodomer (Greg. Tur., ni 18). 2 Beow. 2618 f.: no ymb '5a faeh%e spraec, \>eah ^e he his bro^or beam abredwade. Many scholars here understand 'Safaeh^e to mean not the encounter between Eanmund and Weohstan, but the hostility (vendetta) which devolved upon Onela as Eanmund's kinsman ; but I think the idea is rather that of * bloodguiltiness ' (towards Onela) incurred by Weohstan. Eanmund was the son of Ohthere, Onela's brother. 348 SOCIETY IN THE HEROIC AGE [CHAP. to primitive tribal custom he ought to have taken vengeance upon his knight. It is clear then that primitive custom was breaking down even in countries far removed from contact with Christianity and Roman civilisation. We cannot tell indeed how far the change was general, since our knowledge is practically limited to the princely families. It is by no means unlikely that the lower strata of society were more conservative in many respects. The principle which had now become dominant, at least in the higher ranks, was that of personal allegiance. This principle was of course by no means new. Even in Tacitus' works we hear of the comites who lived and fought in their lord's service and thought it a disgrace to survive his death. In the Heroic Age however it is probable that among the more northern peoples every man, except the king himself, had a lord. In the Anglo-Saxon laws the lord shares with the kindred the duty of protecting his men, and when one of them is slain he receives a special payment (the manboi) when the wergeld was paid to the relatives. Also, when any of his men die, at all events a man of the higher classes, he is entitled to the heriot, i.e. the arms of the dead man, which in theory at least the latter had originally received from him^ But in the poems, as is natural, we hear most frequently of the knights who formed the courts and retinues of kings and princes. As a summary of the services rendered by such persons to their lord, Tacitus' brief description {Germ. 13) still holds good : their presence gave him dignity in time of peace and protection in war. They dwelt and served him at his court and joined him in hunting and other amusements, while he rewarded their services with gifts of treasure and arms. In the descrip- tions of kings which we meet with in the poems there is no characteristic — not even personal bravery — which receives more commendation than that of generosity to their followers. In return the knight was expected to give up to his lord whatever ^ It may be observed that in Beow. 452 ff. the hero requests the Danish king to send his mail-coat to Hygelac, if he should be killed by Grendel. This mail-coat (described as Weland's handiwork) is said to have belonged formerly to Hrethel, Hygelac's father (Beowulfs grandfather). XVI] SOCIETY IN THE HEROIC AGE 349 he gained by his own exploits — just as Beowulf renders up to Hygelac and his queen the valuable gifts which he had received from the king of the Danes. As an instance of personal devo- tion in time of war we may cite the surrender of Chonodomarius' retinue at the battle of Strassburg — an incident to which we have already referred (p. 340). So also in the various accounts of the fall of Hrolfr Kraki given by Scandinavian authorities the king's knights are said to have perished to a man. The same spirit survived in England in later times, as we see from the story of Cynewulf's death, when in each of the two en- counters only one member of the defeated party was left alive. It was also thoroughly in the spirit of the Heroic Age that Edwin's knight, Lilla, acted when he threw himself between the king and the assassin and received a mortal wound in so doing. It was customary for the sons of noblemen to enter the king's service at an early age. Beowulf went to Hrethel's court when he was only seven years old ; but this case may have been exceptional, as he was the king's grandson. When they reached manhood' the king was expected to provide them with estates or jurisdiction over land, which would enable them to marry and support a household of their own. Thus Beowulf, after proving his prowess at the Danish court, is presented by Hygelac on his return with seven thousand hides — a considerable pro- vince — together with a residence and a prince's authority. The grant is accompanied by the gift of a sword, signifying that the bond of personal allegiance was still preserved. Beowulf in turn presents his young relative Wiglaf with the dwelling-place of their family and the public rights appertaining thereto. The court minstrels Widsith and Deor receive grants of land from their lords. In two of these cases (those of Wiglaf and Widsith) we are told that the estate had previously been in the possession of the recipient's father; and we may probably assume that such cases were not uncommon. Yet it is plain that such practices must very largely have destroyed the tribal custom of succes- sion — at least in the higher ranks of society. ' In the seventh century it appears to have been customary to make these grants when the recipient was about twenty-four or twenty-five years old ; cf. Bede, Hist, Abb., §§ I, 8 ; Ep. ad Ecgb., § 11. 350 SOCIETY IN THE HEROIC AGE [CHAP. Those who had received grants of land or jurisdiction did not thereupon cease to attend the court. In the English courts of the seventh century Bede distinguishes between the comites, who already held office, and the ministri or milites, who seem in general to have been young knights without such official posi- tions. The Anglo-Saxon version of the Ecclesiastical History translates comes by gesi^ and minister or miles by jyegn. In poetry both these words are of frequent occurrence, though they appear to be used more or less indiscriminately. It should be observed that the word }}egn means properly no more than 'servant^' though (like knight in later times) it came to be specialised, while gesi'6 is almost an exact equivalent of comes"^. In Beowulf however we meet with the same classes under the collective terms geogd^ and duguS, i.e. youths and men of tried valour respectively. To the latter may be assigned such persons as Aeschere, Hrothgar's "confidant and counsellor," who had stood by his side on the battlefield ; but the former class were probably as a rule in the majority. Another characteristic of these retinues, which deserves notice, is the fact that they were not always composed of born subjects of the king. Bede {H. E. Ill 14) says that Oswine, the popular king of Deira, attracted young noblemen to his service from all sides ; and in the Heroic Age such cases appear to have been frequent. Perhaps the most striking case in the poems is that of Weohstan, who took service under the Swedish king Onela and consequently became involved in hostilities against his own nation. It is probably due to the same custom that we find so many Teutonic chieftains serving under the Romans during the Heroic Age. Among them we may men- tion Arbogastes, Stilicho, Ricimer and Odoacer. Most frequently perhaps the men who sought service abroad were those who had either lost their lords or had had to leave their homes through vendetta. Such cases occur frequently in the Anglo-Saxon poems ; we may refer especially to the Wanderer and the 1 We may compare the use of the word sluga in Servian heroic poetry (cf. p. 316) ; its ordinary meaning is ' servant.' * The same word is used in a similar sense in the Langobardic laws ; gasindus (or gasindius), ' Gefolgsmann,' and so zXso gasindium, ' Gefolgschaft ' ; cf. Bruckner, Quellen und Forschungen, Lxxv p. 205. XVI] SOCIETY IN THE HEROIC AGE 35 1 Husband's Message. Further, it appears from the story of Waldhere and Hagena that even hostages were expected to fight for the prince to whom they had been given. In later times we may compare the case of the British hostage who was wounded in the fight following Cynewulf's murder. But there are a number of other stories which seem to indicate that it was at one time a regular custom for young princes to set out from their homes, on reaching manhood, and to seek the court of some foreign king with a view to marrying his daughter and thereby acquiring a share in the sovereignty. Such incidents are of the commonest occurrence in folktales ; and we find them also in works, such as Hervarar Saga and Ynglinga Saga, which claim to be based on genuine tradition. It is in this light too that all northern authorities represent the position of SigurSr at the home of GutJrun. What has been said above applies primarily of course to the more northern peoples. The Goths were early exposed to Christian and Roman influence, and the same is true also of the Burgundians, especially after their settlement in Gaul. The Franks were no doubt less affected at first ; but their customs seem from the beginning to have differed a good deal from those which we have been considering. They too had retinues of warriors {antrustiones or homines in truste regis) attached to the kings by personal service ; but the prevalence of lordship in the lower ranks of society is by no means so clear. The possession of land also seems to have been governed at first by tribal principles and later by that of succession in the male line — without reference to the will of a superior. These differ- ences are doubless connected with certain features which distin- guished the social organisation of the Franks from that of the other Teutonic peoples. Every one of the early Teutonic nations possessed a more or less elaborate social system, with various class gradations. These gradations may be seen most clearly in the sums of money fixed for wergelds, for the compensations fixed for various injuries and insults and for fines ; in some cases also they show them- selves in the relative value attached to oaths. Apart from slaves, who do not come into consideration in these matters, the classes 352 SOCIETY IN THE HEROIC AGE [CHAP. usually met with are those of nobles, freemen and freedmen. Sometimes however a class is subdivided; sometimes again one class is wanting altogether. Thus in the Anglo-Saxon king- doms, except Kent, there were two grades of nobility, apparently landowning and landless, while freedmen did not form a distinct class. Among the Franks on the other hand we find no noble class. In general the freeman's wergeld is about double that of the freedman, while that of the noble is twice or thrice as great as the former ; and the other payments usually follow more or less the same proportion. The actual sums fixed in the various laws differ greatly in each case, owing to the employment of different systems of currency. But it may be regarded as extremely probable that the normal wergeld of the freeman was originally a hundred head of cattle. Some nations, further, had special wergelds for certain high officials. Among the Franks persons in triiste regis had threefold wergelds, and the same applied to the ordinary freeman when engaged in military service. In England the existence of special official wergelds is uncertain, at least before the great Danish invasion, though such persons were entitled to higher compensations in other respects. But in this country members of the royal families had wergelds six times as great as those of the higher class of nobles. All the above classes (excluding officials of course) seem to have been as a rule hereditary. In some nations indeed the descendants of freedmen did become freemen. But it is scarcely probable that this class everywhere consisted only of manu- mitted slaves or their offspring. Sometimes we find the terms litus, latus, lazzus {laet in the Kentish laws) in place of libertus ; and there is good reason for believing that this class was largely derived from subject populations. Its numbers seem to have been very large. As to the numbers of the nobility there was apparently great difference between one nation and another. Among the Bavarians it consisted only of six families, including that of the duke, whereas in England it appears to have formed a considerable element in the population. The term applied to it here was gesi^cund, i.e. o( gesi^ origin (cf. p. 350), which indi- cates clearly a hereditary official or rather military class. Indeed XVI] SOCIETY IN THE HEROIC AGE 353 the evidence seems to show that the population consisted of two clearly defined classes, which we may describe as military and agricultural, and that all serious fighting was left to the former. This is another feature in which Anglo-Saxon custom differed from that of the Franks, whose armies in the sixth century appear to have been of a more truly national or even tribal character. With regard to the other nations we have less information ; but it is probable that the military organisation of the Danes and other Baltic peoples approximated more nearly to the English type. In Homeric society we find the same forces operative as in that of the Teutonic Heroic Age. The duty of protecting or avenging one's relatives is frequently mentioned. Thus in Od. XVI 97 f. the disguised Odysseus says to Telemachos : " Hast thou fault to find with thy brethren, for it is in them that a man trusts to do battle, even if a great quarrel takes place?" Telemachos replies (v. 115 ff.): "Nor have I fault to find with my brethren, in whom a man trusts to do battle, even if a great quarrel takes place. For our family has been reduced to one man by the son of Cronos, as I will tell thee. Arceisios begat one only son Laertes, and Odysseus again was the only son begotten by his father ; but Odysseus begat me only and left me in his palace without profit to himself. Hence there are now innumerable enemies in our house." The duty of vengeance is clearly recognised by Nestor in Od. Ill 196 ff. : " How good a thing it is for even a dead man's child to survive ! For he (Orestes) also took vengeance on his father's slayer, the crafty Aigisthos, who killed his famous father." It was in order to escape such a fate that Theoclymenos besought Telemachos to take him on his ship {ib. XV 272 ff.) : " I also have left my country, having killed a man of my own tribe (or 'people'). And in Argos, the pastureland of horses, he had many brethren and kinsmen who hold great authority among the Achaeans. I have taken to flight and so evaded death and black fate at their hands ; for it is still my lot to wander among men. Now take me on thy ship, since I have come to thee as a fugitive and suppliant, lest they kill me 354 SOCIETY IN THE HEROIC AGE [CHAP. outright ; for I am sure they are in pursuit." We are reminded here of the story of Ecgtheo, the father of Beowulf, who fled for protection to the Danish king Hrothgar owing to a similar cause (Beow. 459 ff.). Another case of such exile occurs in II. XIII 695 ff. (XV 334 ff.) : " He (Medon) dwelt in Phylace, away from his own fatherland ; for he had slain a man, the kinsman of his stepmother Eriopis, whom Oileus had to wife." We may compare also II. xv 430 ff. where Hector slays "Lycophron, the son of Master, the Cytherian squire of Aias, who dwelt with him ; for he had slain a man in divine Cythera." Among the Homeric Greeks, as in northern Europe, com- pensation for manslaughter could be made to the dead man's relatives. Thus in II. IX 632 ff. Aias says to Achilles: "And yet one accepts compensation from a man who has slain one's brother or for the death of a son. Hence it comes to pass that the one, when he has paid a large compensation, remains in his own land, while the other, after he has accepted the com- pensation, restrains his feelings and his proud spirit." Again, in the description of Achilles' shield we find a scene {ib. XVIII 497 ff.) representing a dispute over the payment of a wergeld. " The folk were gathered in the assembly place ; for there a strife was arisen, two men striving about the blood-price of a man slain ; the one claimed to pay full atonement, expounding to the people, but the other denied him and would take naught ; and both were fain to receive arbitrament at the hand of a daysman \" In this case the transaction takes place before certain elders, one of whom is to receive a payment of two talents, apparently as a reward for bringing about an agreement. There is no mention of any payment to the king^ From the passage relating to Theoclymenos quoted above (v. 273 : TToWot Kaa-iyvTjTOi re erai re) we may probably infer that the duty of vengeance extended beyond the brothers of the slain man ; and evidence to the same effect is given by the story of Tlepolemos (II. II 661 ff.). But it is not at all clear how many degrees of relationship were either involved in this duty or entitled to compensation. Indeed the poems give us very little ^ Quoted from the translation by Lang, Leaf and Myers. ^ The interpretation of Hesiod, fV. and D. 38 f., need not be discussed here. XVI] SOCIETY IN THE HEROIC AGE 355 information regarding the character of the kindred. Patronymic forms such as 'Ar/aetSr;?^ are very common and correspond in use to Anglo-Saxon forms such as Hre^ling, Wonrediyig. But they are almost always used of individuals or of a pair of brothers ('ArpelSa, Soiol 'iTnraalSat). Collective names for families, such as 'HpaKXetBat, UeXoirlSac, seem not to occur in the Homeric poems^ Indeed the patronymic is nearly always derived from the name of the father. Cases where they are taken from the grandfather's name, e.g. AtWi'ST;? for Achilles, appear to be quite exceptional. This is a feature in which Homeric usage differs not only from Teutonic but also, still more, from that of later times in Greece, where we frequently find families or kindreds bearing patronymic names derived from a remote ancestor. As examples we may mention the Aigeidai at Sparta and the Philaidai at Athens. But the difference between Homeric and later usage in this respect does not seem to be one of nomenclature only. At Athens we find later an elaborate system of ' tribes ' {(f>v\ac), phratries or 'clans' ((fiparpat) and 'kindreds' (yepr)), of which the last at all events were supposed to rest on a basis of blood- relationship, involving common religious rites. Divisions of a more or less similar type seem to have existed in the other Greek states. In the Homeric poems however we find extremely little evidence for anything of this kind. The clearest case is in II. II 362 f , where Nestor instructs Agamemnon as follows : " Divide thy men according to tribes and clans, Agamemnon, that clan may render succour to clan and tribe to tribe^" In the battle scenes we hear little of any such organisation, though this may be due to the fact that attention is entirely concen- trated upon the leaders. But it is worth noting that the word * The other types (e.g. U-riXeluv, TeXa/Juivioi) are less frequent. * KaSfieioi, Kadfielwves are at best very dubious examples, for Cadmos is probably to be regarded as an eponymous national hero, like Dardanos. ^ Kpiv' dvSpai /card. p-f^pi) v\ois. Cf. Tacitus, Germ. 7 : quodque praecipuum fortiiudinis incitanienttim est, non casus nee fortuita conglobatio turmam aut cuneum facii sedfamiliae et propinquitates. But the context shows that the conditions here are of a totally different character from those in the Iliad. 356 SOCIETY IN THE HEROIC AGE [CHAP- (fypijTpT] occurs only in this passage'. Again, (f>v\ri does not occur at all, except possibly in the form KaTa. cit., p. 91) has been inferred from the mutinous behaviour of a ship's crew. XVIlJ GOVERNMENT IN THE HEROIC AGE 383 when the assembly should be called, and indeed whether this body had much in common with the constitutionally regulated assemblies of historical times^ It seems rather to be a more or less fortuitous gathering called together on the spot by criers when the king wishes to bring something before the notice of the public^. In Od. Ill 5 ff. we certainly do hear of a great public gathering — indeed we may probably say a national gathering — of a kind which can only have taken place at definitely fixed times. But it is clear that this was essentially a religious festivals Such gatherings may of course have been used for political purposes, as in the North ; but we have no information on this subject. The Achaean 'council of elders' {^ovXtj yepovrcov) in the Iliad seems to be a body of quite as informal character as the assembly. On several occasions Agamemnon calls together a small number of princes, namely Nestor, Idomeneus, the two Aiantes, Diomedes and Odysseus, together with his brother Menelaos. This number of course forms only a small proportion of the leading men in the army. Occasionally however we find ^ It cannot fairly be argued from Od. II 192 f. that the assembly (apart from the king) has a right to impose fines, for the suitors here are relying not upon any * constitutional ' rights but on force majeure. It is to be remembered too that Eurymachos appears to have designs upon the throne (cf. p. 358 f.). ^ It has been suggested that the true name for such a gathering was dduKos (Owkos) and that this was something different from the dyop-^ (cf. Fanta, o/>. cit., p. 77); but the evidence for such a distinction is very far from convincing. We may refer to such passages as Od. XI i 318 and, more especially, to v 3 (BGjKbvbe), which is clearly parallel to II. xx 4 (dYopTjj'Se). Cf. Finsler, N.Jakrb., xiii 327. ^ In the Hymn to the Delian Apollo, v. 146 ff., mention is made of a festal gathering of lonians at Delos, apparently on a considerable scale. Similar gatherings may have been in existence quite as early, or even earlier, in other parts of Greece. For the festival at Pylos however much better parallels are to be found in the great religious gatherings which took place every nine years at Leire and Upsala, the old Danish and Swedish capitals. Cf. Thietmar of Merseburg, Chron. 19: est unus in his partibus locus... Lederun nomine... ubi post nouem annos, mense lanuario . . .omnes conuenerunt et ibi diis suismet XCIX homines et iotidem equos cum canibus et gallis pro accipitribus oblatis immolant. And Adam of Bremen, iv 27 : solet quoque post nouem annos communis omniuin Sueoniae prouintiarum sollempnitas in Ubsola celebrari. ad quam uidelicet sollempnitatem nulli praestatur immunitas. reges et populi omnes et singuli sua dona transmittunt ad Ubsolam. It does not appear however that on these occasions — in contrast with the festival at Pylos — any of the victims were eaten. In this respect they are probably to be compared rather with the great quadrennial sacrifices of the Gauls ; cf. Diodoros, v. 32. 384 GOVERNMENT IN THE HEROIC AGE [CHAP. Others summoned, such as Meges and even Thrasymedes and Meriones, who are not the chiefs of contingents. The council of an expeditionary army however is an exceptional case. On the Trojan side we hear of a number of 8r)/j,oyepovT€^ with Priam (II. HI 146 ff.), seven of whom are named. Three of them are brothers of the king, while others are fathers of the most dis- tinguished Trojan warriors. They are described as eloquent orators, but no account is given of their deliberations. In the Odyssey references to councils are very rare. No mention is made of such a body in Ithaca. In Scheria however Alcinoos has twelve kings under him (VIII 390 f.), who clearly form his council and are to be identified with the ' leaders and rulers of the Phaeacians ' {^atijKcov r)&)«?)6yrs), etc. XVIII] RELIGION IN THE HEROIC AGE 413 Heroic Age and among many of the Teutonic peoples. It is perfectly true that the notices of Teutonic religion contained in Tacitus' works convey the impression that religion was re- garded as a very serious matter and that the general attitude towards the gods was highly reverential. The same impression is conveyed by Alcuin's account of Fositesland ; and probably no one will deny that the euthanasia of the Heruli was based upon a very real conception of immortality. But to compare such records with the poetry of the Edda would manifestly be absurd. For analogies to them we must turn to notices relating to actual religion, and here we shall find evidence that the people of the Viking Age were no less religious than those of earlier times. We may instance the reverence shown by Th6rolfr of Mostr to his holy hill and Earl Haakon's devotion to ThorgerSr HolgabriiSr. On the other hand the attitude towards the gods shown in the Edda finds an exact analogy in the only record of ' theo- logical' poetry which has survived from the Heroic Age. In the Langobardic story (cf p. 115) the anthropomorphisation of the deities is already complete ; and the chief god^ is duped by his wife. We could scarcely wish for a better parallel to the account given in the introduction to Grimnismdl. In view of this story it is scarcely possible to doubt that familiarity, not to say levity, in the treatment of the gods characterised the poetry of the Heroic Age, just as much as that of the Viking Age. It would be well to hesitate however before assuming that the gods of Tacitus' time were treated in the same way. His account shows that Teutonic theology had then passed beyond the purely tribal stage, and that certain deities were worshipped by a number of peoples, if not universally. But it does not suggest the existence of a highly anthropomorphic conception of the gods. Further we have to bear in mind that Tacitus is separated only by a century and a half from Caesar. The account of German religion given by the latter {B. Gall. VI 21) is difficult to account for by any explanation. But unless we are to believe that Caesar was thoroughly imposed upon we ^ It will be observed that here (as commonly in the Edda) Wodan's character as god of the dead (slain) is entirely lost sight of. 414 RELIGION IN THE HEROIC AGE [CHAP. must conclude that nothing in the nature of a developed poly- theism can have existed in his day. To the theology of the Heroic Age his account of the Gaulish gods {ib. VI 17) would be far more applicable than what he says regarding the worship of the Germans. The question we have been discussing appears to throw some light upon the rapidity with which most of the Teutonic peoples accepted Christianity. The facts which we know with regard to the conversion are as follows: (i) that it almost invariably began in the king's court ; (2) that violent opposition was offered only in kingless communities, as among the Old Saxons, or in defiance of the king's authority, as in Norway; (3) that after the conversion the gods (in general) disappear at once and for good ; (4) that magical practices and the belief in spirits and even in certain female agricultural deities (' Erce,' Holda, Berhta, etc.) lasted among the country people for many centuries. From (4) we may probably infer that the religion of the country people was chiefly animistic — similar no doubt to what we find in Iceland, with the exception that we have little evidence for the cult of the thunder-god. Again, the explanation of (3) hangs together with (i); for the statements of ecclesiastical writers render it clear that the religion of the courts was essentially theistic. But it is plain from the dis- cussion in the Northumbrian council recorded by Bede (^H. E. II 13) — the only discussion of this kind of which we have any detailed account — that here at least this religion retained little vital forced This fact is fully explained if, as I have endeavoured to point out, theology had largely passed from the realm of dogma into that of poetry. The conclusion then to which we are brought is that Teutonic religion, at all events in the courts, underwent a profound change in the course of the Heroic Age. It is to be observed that in the earlier part of that age — as in the earlier part of the Viking Age — we find, especially among the more northern peoples, ^ We may note especially the two speeches of Coifi, the chief priest : Nihil omnino uirtutis habet, nihil utilitalis religio ilia quam hucusque tenuimus — lam olim inlellexeram nihil esse quod colebamus ; quia uidelicet quanta studiosius in eo cultu quaerebam, tanto minus inueniebam. XVIIl] RELIGION IN THE HEROIC AGE 415 a fanatical devotion to warfare for its own sake, accompanied by lust for destruction and apparently also by a vivid conception of a life hereafter. In the latter part of the Heroic Age these phenomena disappear, except among the Heruli who, according to Procopius, differed from all the rest of mankind. The ideal which the princes of the later period set before themselves may be gathered both from Beowulf and from Roman authorities ; it was to enjoy wealth and splendour in this life and to have their fame celebrated by future generations. For their attitude towards a future life the speech of the Northumbrian councillor recorded by Bede {I.e.) may probably be regarded as typical'. In a work such as this it is scarcely necessary to give even a brief summary of the characteristics of Greek religion, since the main outlines of the subject are probably much more familiar than those of even the later religion of the North. At the same time the amount of information which has been preserved is so great and the unsolved problems presented by the subject so numerous that it is clearly better left in the hands of experts, I shall attempt no more therefore than to call attention to the salient points in which the religion of the Homeric poems differs from that of later times and to the chief characteristics in which the former resembles or differs from the religion of the Teutonic Heroic Age and the Viking Age. The various objects of worship recognised in Greece belong in general to much the same categories as those which we have noticed above. They may be classified roughly as gods, genii locorum and manes. In the last class we may perhaps include the ' heroes,' though the position occupied by them is somewhat peculiar. They were for the most part characters of the Heroic Age, and sometimes we find the cult of the same hero recognised in a number of different states. In general the worship of the gods took a different form from that paid to the manes and heroes, though occasionally the cult of a deity seems to have been associated with, or superimposed upon, that of a hero. ^ J.ta haec uita hominum ad modicum apparet; quid auteni sequatur, quidue praecesserit, prorsus ignoramus. 4l6 RELIGION IN THE HEROIC AGE [CHAP. In the Homeric poems the gods figure much more prominently than the other classes. The most frequently mentioned of them are Zeus, Hera, Poseidon, Athene, Apollo, Artemis, Hephaistos, Aphrodite, Ares and Hermes. Zeus, as head of the divine com- munity, corresponds to Othin (Woden); but in other respects there is little resemblance between the two. As god of the thunder his affinities are rather with Thor. Hera, as wife of the chief god, may be compared with Frigg, while Poseidon, as god of the sea, has an element in common with NiorSr; but he is also an earthquake god, which the latter is not. His characterisation also is much more clearly marked, Apollo and Artemis, as a pair of young deities, brother and sister, with certain characteristics in common, bear some resemblance to Frey and Freyia ; but the sexual element, so prominent in the Northern deities, is wanting in the Homeric poems. On the whole the translator of the Icelandic New Testament seems to have been happily inspired in rendering Diana (Artemis) by Gefion}, while Freyia's true counterpart is rather to be found in Aphroditel Hephaistos, the smith of the divine community, has no Northern god corresponding to him ; as the maker of heroes' armour and other metal objects he plays the same part as Weland. Ares, as god of war, has an element in common with Tyr. The Homeric Hermes bears no resemblance to Othin ; his duties are to a certain extent discharged by the Valkyries in Northern mythology. Athene is a character totally foreign to Northern theology. All the above deities, together with a number of others less important, form a regularly organised community, like the Aesir. Their home is located on Mount Olympos in the north of Thessaly — a conception probably more primitive than AsgarSr, Svhich is never represented as a place known to the human race. But, though Olympos is the home of the gods collectively, most of them (like the Northern deities) have also one or more 1 Artemis in the Homeric poems seems to be chiefly a women's deity (cf. Od. xx 59 ff.). The same may have been the case with Gefion; cf. Gylfaginning, cap. 35: "She is a maiden, and those who die in maidenhood serve her." 2 In origin however Freyia, the daughter of Nior'Sr (Nerthus), corresponds probably rather to Persephone, the daughter of Demeter ; cf. p. 400. XVIIl] RELIGION IN THE HEROIC AGE 417 dwelling-places of their own, often in distant localities. Thus Poseidon's home is at Aigai and Apollo's at Delphoi, though he is also connected with several localities on the eastern side of the Aegean — Chryse, Cille, Tenedos, etc. Athene has a home at Athens, while Hephaistos is connected with Lemnos, Ares with Thrace and Aphrodite with Paphos in Cyprus. Zeus himself, apart from Olympos, has abodes at Dodona and Ida. The belief that the gods had homes of their own in various localities is clearly to be taken in connection with the fact that they are said to have sanctuaries in the same places. Thus Aigai is mentioned, together with Helice, as a place where sacrifices were offered to Poseidon (II. VIII 203 f ). Zeus' home at Dodona is his sanctuary {ib., xvi 233 ff.), and it is to her sanctuary at Paphos that Aphrodite goes (Od. VIII 362 f ). We need not doubt therefore that, as in later times, the cults of the various deities were largely of a local character. It is a different question of course whether the cults — or rather the deities them- selves — were of local or tribal origin ; but in certain cases such an explanation appears to be by no means improbable. Many scholars believe that Ares was originally a Thracian deity; and the fact that he is represented as the father of Boeotian heroes can scarcely be regarded as conclusive evidence to the contrary. In II. IV 5 1 f. Hera states that three cities, Argos, Mycenae and Sparta, are specially dear to her; and there is little evidence that her cult was ever prominent in any other part of the Greek mainland. If the same explanation is true of such deities as Apollo and Aphrodite we must suppose either that their cults have spread from one locality to another or that deities belonging to different localities have been identified. The local origin of river gods, such as Spercheios and Scamandros, is of course clear enough ; but these are little more than genii locorum. But the really important feature in Homeric theology is that the various deities, whether of local origin or not, are all brought together as members of one community, or rather family, and that as such they are represented as meeting with universal recognition. Indeed, they are not regarded even as peculiar to the Greek race ; the Trojans recognise the same divinities, and Poseidon goes to receive sacrifices even from the Ethiopians. 27 41 8 RELIGION IN THE HEROIC AGE [CHAP. A similar, though less important, community is that of the marine deities, one of whom, Thetis, figures prominently in the Iliad. The spirit in which the gods are treated is in general very similar to what we find in the Edda, and in no way more reverential. Sometimes the treatment is humorous, as in Thrymskvi-ga ; sometimes again the deities are represented in a very unfavourable light. For the trick played upon Wodan by Fria in the Langobardic story (cf p. 115) we have an interesting parallel in II. XIV 292 — 353, where Hera distracts Zeus' attention from the war and coaxes him to sleep. But this is only one of a number of incidents which give a generally unpleasant picture of the domestic life of this pair, much more so than is the case with Othin and Frigg in the Edda. Again, the story of Ares and Aphrodite, which forms the subject of Demodocos' lay (Od. VIII 266 — 366), is very much what we should expect to find in a poem dealing at length with the scandalous charges brought against the goddesses in Lokasenna. Ares' speech in v. 292 ff. — as also Hermes' remarks in v. 339 ff. — may be compared with those of Frey in Skirnismal. In the Iliad the same two deities are more than once treated contemptuously. The same spirit is shown in the treatment of the relations of gods with men. It is frequently recognised that deities ought to show gratitude to their worshippers for the sacrifices offered to them, as Freyia does in Hyndluli6S, though at the same time they are at liberty to refuse a petition, as in II. VI 311 — a case which may be compared with Frey's conduct in Viga-Glums Saga (cf. p. 253). In the Iliad we find several deities taking an active interest in the fortunes of the war — Poseidon, Hera and Athene on the side of the Achaeans, and Apollo, Ares and Aphrodite pn that of the Trojans. Poseidon more than once comes to rouse the Achaeans, when he thinks they are becoming slack. Apollo and Athene interfere in the struggle in a manner which strikes the reader as unfair ; indeed the latter is frequently guilty of flagrantly mean and dishonourable conduct towards her opponents. The attitude of the heroes themselves to the gods is quite in keeping with the way in which the latter treat them, Achilles openly abuses Apollo for deceiving him XVIII] RELIGION IN THE HEROIC AGE 419 (II. XXII 1 5 ff.) ; and Diomedes attacks and wounds both Ares and Aphrodite, but only after assurances or assistance from Athene. In this respect the Homeric princes are not quite so bold as the hero of Biarkamal (cf p. 402 f ). But it is by no means only towards communities that the deities display their favour or hatred. Indeed their attitude towards the contending forces at Troy seems to be largely determined by their relations with certain individuals, especially Helen and Paris. In the Odyssey Poseidon is represented as persecuting Odysseus, one of the Achaean leaders, for many years on account of the injury done by him to Polyphemos. On the other hand both this hero and Diomedes enjoy the special favour of Athene. It is worth noting that Diomedes appears to have inherited this favour from his father (cf II. v 800 fif.), and the same may be true in the case of Odysseus (cf Od. XXIV 367 ff, 517 ff.) ; at all events it is continued towards his son. In neither of these cases is any hint given of conjugal relations, such as we find in the North ; indeed the sexual element seems to be practically absent from Athene. Elsewhere however such relations are clearly involved, e.g. in the cases of Aineias and Achilles, who are the sons of Aphrodite and Thetis respectively. In the Odyssey (v 119 ff.) the principle is stated plainly; and the hero himself has conjugal relations with Circe and Calypso, both of whom are described as goddesses. Of other mythical beings the genii locorum are perhaps the most prominent. The distinction between them and the class with which we have been dealing is by no means so clearly drawn as in the north of Europe (cf p. 407 f ). Many of them are even described as gods (deoi). River gods, such as Spercheios and Axios, are represented as the fathers or ancestors of several heroes. Among other, less important, beings of the same type we may mention the nymphs^ to whom certain caves and springs were sacred (e.g. Od. XIII 347 fif, XVII 205 ff, 240 ff). References to chthonic deities are not very frequent. The Erinyes are mentioned several times, especially in curses ; but ^ Even beings such as these are sometimes affiliated to Zeus ; cf. II. vi 420. 420 RELIGION IN THE HEROIC AGE [CHAP. they are not individualised, though the singular is occasionally used. The most important person in the under-world appears to be the queen, Persephoneia ; but she is never actually brought upon the scene. Hades himself is seldom more than a name. The sacrifices to the gods mentioned in the Homeric poems are as a rule similar to the sacrificial feasts of which we hear in the Norths A portion of the victims, usually bulls or rams, was offered to the deity, while the rest was consumed by the worshippers. On the other hand we hear occasionally of victims which were given wholly to the gods. Such appears to have been the case with the victims sacrificed on the occasion of a solemn oath ; thus in II. XIX 267 f. the boar is thrown into the sea^ It is perhaps worth noting that both here and in III 276 fif. the oath refers to chthonic deities', as well as to Zeus, the Sun and Earth. Victims sacrificed to the dead likewise seem to be offered entire (e.g. Od. XI 44 ff.; II. xxiii 166 ff.). Moreover it is only in connection with funeral rites that we hear of human sacrifices, namely when Achilles puts to death twelve Trojan youths at the pyre of Patroclos {ib., 175 f.). Horses and dogs are also sacrificed on this occasion, as well as oxen and sheep, and the whole scene is in accordance with Northern custom. No mention is made in the poems of sacrifices such as that of Polyxene, which form so favourite a theme with the dramatists ; but in view of the Northern evidence^ we are entitled to doubt whether they are altogether inventions of later times. 1 So also with the ceremonial drinkings, which in both sets of records form so prominent a feature in the life of human and divine communities alike. Among the Greeks the libation corresponds to the Northern 'toast' (full) in honour of the gods. A good example may be found in Saga Hakonar Go'Sa, cap. i6. 2 Reference may be made to the boar which was sacrificed to Frey (or Freyia) in Hervarar Saga, cap. lo, and on which oaths were sworn. It is not made clear however what was done with the body of the boar. ^ With the punishment of perjurers by the Erinyes in II. xix 259 f. we may compare what is said regarding the fate of such persons in Volusp^, str. 39. * Reference maybe made to Procopius' account of the Heruli (cf. p. 411) and more especially to the stories of Ibn Dustah and Ibn Fadhlan, quoted by Thomsen, The Relations between Ancient Russia and Scandinavia, etc., pp. 30 f., 34. Native records preserve only somewhat vague traditions relating to such a custom ; cf. The Cult of Otkin, p. 4I f. XVIII] RELIGION IN THE HEROIC AGE 421 The theology of the Homeric poems received formal recog- nition in Greece down to the acceptance of Christianity ; but there is no evidence that the system as a whole possessed any vital force even in the earliest times of which we have record. From the sixth century onwards poets and philosophers began to regard Zeus as much more than the chief of the gods ; but even in popular religion it appears that each state honoured certain deities, while the rest were largely or altogether neglected. Thus at Athens more prominence seems to have been given to Athene, and perhaps also to Poseidon, than to Zeus, while Hera was predominant at Argos and Samos, and Apollo at Delphoi, etc. This is a feature for which the Homeric poems themselves give evidence, as we have seen, and there is no reason whatever for supposing it to be of later growth. Sometimes too we find prominence given to deities, such as Demeter and Dionysos, who seem to be of little consequence in the Homeric poems, while other cults, such as that of Serapis, were introduced from abroad in comparatively late times. Chthonic deities are more prominent in later literature than in the Homeric poems ; and many authors describe the worship paid to them down to a comparatively late period. Yet the rites seem to have been of a primitive character. But perhaps the most striking element in the religion of classical Greece was the worship of ' heroes.' Every city possessed shrines {rjputa), at which sacrifices were offered to heroes with rites similar in general to those used in the worship of chthonic deities. Originally these heroes seem to have been local persons and the shrines their tombs ; but characters prominent in heroic poetry sometimes received worship in many different states. In the Homeric poems there is no evidence for this hero- worship. Indeed these poems contain few traces of a cult of the manes at all, except in funeral ceremonies and in connection with the necromantic sacrifice of Odysseus. This fact has been connected with the Homeric doctrine of immortality^ and in view of the Northern evidence (cf p. 397 fif ) there can be little doubt that the true explanation is to be found herein. The ^ Cf. especially Ridgeway, The Early Age of Greece, p. 512 ff. 422 RELIGION IN THE HEROIC AGE [CHAP. method used in the disposal of the dead is cremation, and, as in the North, it was believed that this sent the spirit away from the body to a place of the dead — not a separate place for each particular family or community but a common home for the souls of the whole Greek race. In II. XXIII 65 fif. the spirit of Patroclos comes and exhorts Achilles to pay him the last rites : "Never again shall I return from Hades, when ye have allotted me the due of fire" (v. 75 f.). So also when Odysseus visits the home of Hades the first spirit he meets is that of his follower Elpenor, who reproaches him with not attending to his obsequies and begs him, when he returns to Circe's island, to "burn him up with his arms, all that he possesses." The idea is clearly the same as in Ynglinga Saga, cap. 8. The honourable way therefore to treat a fallen foe is not to strip him but to burn him with his armour, as in II. VI 417 ff.^ Even at the beginning of the historical period this belief seems not to have entirely died out. According to the story told by Herodotus (v. 92), when Periandros sent to consult the spirit of his wife Melissa, she complained that the clothes which he had given her at her funeral were of no use because they had not been burnt. It has already been mentioned (p. 261) that this story seems to show that at one time the home of the dead was located in a definite, though probably not very well known, region ; and the vague indications given as to the hero's wanderings in the Odyssey can hardly be regarded as evidence to the contrary. If this is correct the home of Hades is in one respect probably a more primitive conception than Valhalla, just as Olympos is more primitive than AsgarSr. A reasonable explanation would be offered if we had evidence that part of the population of Greece was believed to have come from that region ^ An idea of this kind was certainly in Snorri's mind when he wrote the Ynglinga Saga ; but unfortunately the account which he has given is obviously, at least to a large extent, of scholastic origin. ^ Reference may be made to the story of Haraldr Hilditonn and Sigur^r Hringr; cf. The Cult of Othin, p. 22 f. ^ The existence of a tradition to this effect is perhaps implied by Aristotle, Meteorolog., I 14, 21 f. ; cf. p. 437, note. XVIII] RELIGION IN THE HEROIC AGE 423 In Other respects however the Homeric conception of im- mortaHty appears to be less simple than the doctrine of Valhalla, as we find it in certain Northern records. The use of the verbs dcLTrreLv and Tap^vcrovat is probably capable of a different ex- planation ; but there are certain other words and expressions, e.g. KarrjXdev, evepoi, Zei"? Kara')(jd6vioro uiatico cues et boues et alia animantia exhibebant. seruos etiam et aftcillas cum rebus aliis ipsis assignantes una cum mortuo et rebus aliis incendebant, credentes sic ituensos ad quandam uiuorum regionem feliciter pertingere et ibidem cum pecorum et seruorum sic ob gratiam domini combustorutn miiltitudine felicitatis et uitae temporalis patriam inuenire. This record dates from the thirteenth century (probably about 1260) ; so that Northern influence is not impossible. 424 RELIGION IN THE HEROIC AGE [CHAP. determine what may be regarded as characteristic of the Heroic Age. In this category we may probably include the following features. 1. The religion was predominantly a worship of gods, rather than of spirits. Herein lies the chief contrast with the religion of later times in both regions^ In Classical Greece chthonic worship and hero worship seem on the whole to be more promi- nent than that of the gods. The same is true of Scandinavian countries in the latest heathen period, though Thor is still prominent and hero worship is scarcely distinguished from manes worship. In Germany and England, where Christianity was adopted during the Heroic Age or soon afterwards, the gods disappear at once, while forms of chthonic worship survive for centuries. 2. The same gods were, to a large extent at least, recognised everywhere. Whether by borrowing or by identification of cults they had ceased to be merely tribal deities. How far back this feature goes in Greece we cannot tell. Among the Teutonic peoples we can trace it in part back to Tacitus' time ; but it was probably intensified during the Heroic Age. 3. The conception of the gods was definitely anthropo- morphic. For the Teutonic Heroic Age this is made clear by the Langobardic story. There is no absolutely conclusive proof that the gods in general were regarded as forming a regularly organised community, as in the Edda and the Homeric poems ; but all the evidence which we have (cf p. 407 f ) points in tliis direction. 4. The relations between gods and human beings are of a somewhat peculiar character both in Teutonic poetry and in the Homeric poems ; but they are almost identical in the two cases. The gods are not treated with any very great reverence. The conduct attributed to them is not unfrequently repre- hensible, their purposes can often be thwarted by the help of other gods, and the bravest warriors are sometimes even ready * This contrast is all the more noteworthy in view of the fact that the two chief ' heroic ' deities, Woden (Othin) and Zeus, are of essentially different origin. The affinities of the former lie rather with Hades— not as a chthonic being, properly speaking, but as lord of the spirit world. XVIII] RELIGION IN THE HEROIC AGE 42$ to attack them openly. Yet the human and the divine are not confused ; a man is not a god, though many heroes, both Teutonic and Greek, are sprung from gods. It should be observed that the deity from whom most English princes claimed descent is Woden, a universal and not a tribal god. This belief must be regarded as an anti-tribal force. 5. Both in Northern tradition and in the Homeric poems the practice of cremation was associated with the belief in a common home for the souls of the dead. This practice does not appear to have been common even in the earlier part of the Viking Age ; and consequently the tradition probably comes down from the Heroic Age, at which time we know that crema- tion was widely prevalent. The cheerless home of Hades^ differs considerably from Valhalla, though there is some ground for suspecting that the Greeks of the Heroic Age had once cherished a belief endowed with greater vitality. But both con- ceptions possess certain essential features in common, namely the removal of the soul to a distant place — a belief really incompatible with the local worship of heroes or manes — and the fact that this distant place of souls was a universal home and not reserved for the souls of one tribe. This belief again was doubtless an anti-tribal force of considerable importance. Briefly then we may define as the predominant characteristic of heroic religion, both Greek and Teutonic, the subordination of chthonic and tribal cults, which as a rule go together, to the worship of a number of universally recognised and highly an- thropomorphic deities — coupled with the belief in a common and distant land of souls. These characteristics, at all events ^ The characterisation of the home of Hades resembles rather that of the Northern abode of Hel (cf. p. 400, note). There can be little doubt that the latter conception is founded upon a belief of considerable antiquity, though, except in the North, all our information relating to it comes from sources affected by Christian influence. From references in early poetry, as well as from popular belief in later times, it seems probable that the early Teutonic conception of Hell (Goth, halja, etc.) involved something more than a survival of the soul in or about the place of sepulture, though there is nothing to show that it vi^as identical with Valhalla. For the personification of Hell definite evidence is wanting ; yet note should be taken of the fact that such beings as Holda and Berhta (like Freyia and Gefion) were connected with the spirit world. 426 RELIGION IN THE HEROIC AGE [CHAP. among the Teutonic peoples, seem properly to have belonged only to the religion of the royal and military classes. Hence^ when the royal families are converted to a new faith, as in England, or when kingless states grow up, as in Iceland, we find in all cases more or less of a reversion to the more primitive forms of religion. It is on the same principle that I would account for the differences in religion between heroic and historical Greece. XVIII] THE HEROIC AGE OF THE GAULS 427 NOTE VIII. THE SOCIAL, POLITICAL AND RELIGIOUS CHARACTERISTICS OF THE CELTIC AND SLAVONIC HEROIC AGES. In the last three chapters we have confined our attention almost exclusively to the Teutonic and Greek Heroic Ages. It is not to be overlooked however that similar phenomena occur elsewhere. The closest and most interesting parallels, at least in Europe, are probably to be found in the history of the ancient Gauls. It has been mentioned above that Gaulish hterature has entirely perished. We are dependent therefore for our information upon a few scattered references in the works of Greek and Roman writers. These however are sufficient to show that the Gauls possessed a well-known and influential class of pro- fessional minstrels (^ap8ot), whose chief occupation seems to have been the composition of heroic poetry*. Like the minstrels of the Teutonic and Greek Heroic Ages these persons were attached to the courts of kings, and their poems dealt with the praises of living princes as well as with the deeds of heroes of the past I With the disappearance of kingship, early in the first century (B.C.), their standing seems to have been impaired. At all events they are never mentioned by Caesar, who has so much to say about the Druids. It is probable therefore that by this time what may be termed the Gaulish Heroic Age was already at an end. How long it had lasted we cannot tell, since all the stories have perished. We may certainly note however that the accounts of Gaulish life which have come down to us from the time before the nation became Romanised show a most striking resemblance to the conditions described in Teutonic and Greek heroic poetry. The longest of these accounts is the one given by Diodoros (v 26fif.), where their customs are described with a considerable amount of detail. The picture which he ^ Cf. Ammianus Marcellinus, XV 9. 8 (probably from Timagenes) : et Bardi quidem fortia uirorum illustrium facta heroicis composita uersibus cum dulcibus lyrae modulis cantitarunt. Cf. also Diodoros, v 31; Strabo, iv 4. 4; Lucan, Phars. i 447 ff- ^ Cf. Athenaios, VI 49 (quoting Poseidonios) : KeXrot irepidyoi'Tai ned' avrujv Kal iroXe/jLovvTes (Tvn^iwras oDs KoKovffi irapacriTovs. ovtol 5e iyKwfiia avTwv Kal wpbi ddpdovs \iyov(Ttv dvOpibvovs cvveardTas Kal irphs ^KaffTou tQv Kari, /xipos iKelvwv dKpowfxivwi'. TO, 5i aKovanara airrwv el, 'OpicrTTjs, Gu^cttt/s. It is perhaps not without significance that the two latter of these occur in the most important of the Achaean families, while the antiquity of the last is guaranteed by the aspirate (as against Av^a-rai). ^ The extent of these movements may be estimated by the number of names which occur both in the Balkan peninsula and on the east side of the Aegean, e.g. Brygoi (Phryges), Dardanoi, Moisoi (Afysot), Mygdones., Sintoi (Smites), Pelasgoi; and it is not to be overlooked that on the European side most of these names occur in the basins of the Axios and Strymon or the region between them. Attention should be paid also to the common element in place-names, more especially to those (e.g. n^^ayiios) which occur in the islands (Crete, etc.) as well as in the two areas under discussion ; for they seem to indicate that the movements from the Balkan peninsula were not confined to the Asiatic mainland. Even in Asia Minor itself however the Thraco-Phrygian movements may at first have affected a much larger area than that in which languages of this type survived in historical times. Herodotus (l 171) states that on the ground of common ancestry the Carians allowed Lydians and Mysians to use the temple at Mylasa — a privilege which they did not concede to other peoples,^ even to those who spoke the same language as themselves. This passage is usually interpreted as pointing to a traditional religious federation of the indigenous peoples. But it is a serious objection to this view that the Mysians were clearly of European origin. If the Lydian and Carian languages were non-Indo- 45^ THE CAUSES OF THE HEROIC AGE [CHAP. It is by no means impossible that the Shardina were the direct ancestors of the Achaeans. On the other hand they may have belonged to an earlier wave of Greek invasion. Or again they may have been a non-Greek (perhaps Thraco- Phrygian) people whose relations with the Achaeans were rather in the nature of influence, however deeply this may have penetrated. That the Achaeans also were an essentially military people is shown by the tone of Homeric poetry throughout and by the story of the Iliad as a whole, as well as by many inci- dental passages in the Odyssey. We have good reason too for believing that their occupation was in the nature of a military rather than a tribal settlement. That is shown not only by the social and political conditions reflected in the Homeric poems* but also by the fact that, except in two comparatively unim- portant districts, they disappeared after the Heroic Age as completely as the Ostrogoths. The course of our investigations has led us to conclude that there is no reason for regarding the Greek Heroic Age as an exception to the general rule applying to such phenomena. Neither here nor in any other case are we justified in believing that the Heroic Age was a native outgrowth from an ancient and highly developed civilisation. It does not appear that a Heroic Age can arise from such conditions, any more than from conditions which may properly be called primitive. In four of the six cases which we have considered — and we need scarcely hesitate to reckon the Greek case as a fifth — the Heroic Age can be traced back to a similar series of causes. Firstly, we find a long period of 'education,' in which a semi-civilised people has been profoundly affected from without by the in- fltfence of a civilised people. Then a time has come in which the semi-civilised people has attained to a dominant position European, as is commonly believed, they may have been taken over from indigenous peoples such as the Caunioi, whose language was similar to that of the Carians and who seem to have been among the peoples excluded from the temple at Mylasa. The possibility that the Carian language belonged originally to the Caunioi is suggested by Herodotus himself. ^ In particular note may be taken of the absence of national names for the populations of several of the most important kingdoms ; cf. p. 389, note. XIX] THE CAUSES OF THE HEROIC AGE 459 and possessed itself, at least to some extent, of its neighbour's property. The phenomena which we have recognised as characteristic of the Heroic Age appear to be the effects produced upon the semi-civilised people by these conditions. For the exceptional case — that of the Mohammedan Servians — a special explanation has been suggested. Whether this ex- planation be correct or not, I do not mean to assert that the Heroic Age is universally due to the same conditions. They can scarcely hold good for the Irish Heroic Age ; and outside Europe also there are cases, e.g. among the Bantu peoples, of societies which may be called ' heroic ' and yet would probably require a different explanation. All such cases doubt- less postulate conditions so far advanced as to permit the existence of a class of persons who have the opportunity and the ambition to assert their individuality among and above their compatriots. To deal adequately with these cases how- ever would require a greater amount of ethnological knowledge than I possess. I have ventured above to suggest that ' Mars and the Muses ' are necessary for the formation of a Heroic Age. But beyond this I will not attempt to formulate a definition of the elements which constitute a Heroic Age in general. My object has been to call attention to certain common characteristics exhibited by a limited number of epochs in European history. The various Heroic Ages of Europe are usually connected with considerable movements of population. There is some reason for suspecting that this may be true even of the Irish Heroic Age. But such movements do not necessarily produce a Heroic Age. We have no evidence for the existence of a Heroic Age resulting from the great movement of the Slavs into eastern Germany during the fifth and following centuries'. Presumably the antecedent conditions were wanting. So also with the Dorians. It was only the wreckage of the old Aegean civilisation with which they were brought into contact. The ^ It may be observed also that some seven centuries elapsed between the settlement of the Servians on the lower Danube and the beginning of their (first) Heroic Age. This latter period was of course accompanied by an extension of the area occupied by the Servians, though the newly won territories were subsequently lost through the Turkish conquest. 460 THE CAUSES OF THE HEROIC AGE [CHAP. days of mercenary service too, with all its civilising and at the same time denationalising influence, had apparently passed away long before they came to the front. The general direction taken by these movements — though here the Irish case is an exception — was towards the culture lands of the south — i.e. the direction taken by movements of population was the opposite of that taken by movements of culture. The effect of the movements which took place in the Teutonic Heroic Age was to produce a series of inclined or tilted strata of population over a large part of Europe. Thus the Franks formed a ruling aristocracy in Gaul ; but the subjects {coloni, lati) of the Old Saxons were in all probability largely of Prankish blood. Again, the Vandals in Africa were the wealthiest and most luxurious community known to Procopius^ Yet in Genseric's time (cf. p. 369) there still remained a Vandal population in the old home of the nation, of which all traces had disappeared within the next century. Presumably they had been overwhelmed by the surrounding peoples. Similar phenomena are to be found among the Goths and other Teutonic peoples, and in earlier times among the Gauls and Greeks. In the latter case we may cite as an instance the Cynurioi (on the east coast of the Peloponnesos), who were believed to be lonians, though they had been absorbed by the Dorians of Argos. The fact therefore that we hear of no people called Achaeans in Epeiros cannot be held to prove that the Achaeans had never inhabited that region. We hear sometimes from legends of national migrations caused by insufficiency of food. Historical records seem to show that such movements were more frequently due to pressure from neighbouring peoples. But it is greatly to be doubted whether these movements usually involved a total displacement of population. Cases like that of the Vandals indeed indicate that frequently the more enterprising part of the community were the first to move and that the chief impulse came from ^ Vand. II 6: tdvdv yap airduTuv wv ifffj-ev i]fj.eU a^pdraTov fiev to tQv BafdiXwv... ^aXavelois re oi ^v/jLTravres iwexp^vTO is ijfiipav eKacTrfv koI rpairi^y airaaiv €vdrivo6ffr}, bcra di] yfj re Kal ddXaffaa ijdi.bpow 8i ws iirl irXeiarov, Koi 'yirjdiKTjv iadrjTa, ^v vvv "Z-qpLKrtv KoKovaLV, dfi.irexifJ'fi'oi Kal i^Krjvro fxiv aiirCif oi TToWol if irapadeiffots, iibdruv Kal 8iv5pwv ev ^xoi'O'ti k.t.X. XIX] THE CAUSES OF THE HEROIC AGE 461 the attractions offered by the chance of Hving upon the fruits of others' labour, whether in the form of plunder or tribute. So far as our records go back, we find among the Teutonic peoples, as among the Gauls and the early Greeks, a numerous class who prefer the military life to the labour involved in agriculture. Indeed one of the most remarkable features of the Teutonic Heroic Age — and probably of the Gaulish also' — is the ease with which immense hosts of warriors could be gathered for an enterprise of plunder or conquest. It is certain that these hosts were frequently drawn from far and wide. If the national kings would not embark on such enterprises their subjects were ready to embrace the service of neighbouring princes, or even that of distant or alien nations, such as the Romans. It is the existence of this military element which in various epochs of European history and under similar cultural conditions has produced the phenomena comprised under the term ' Heroic Age.' For the special characteristics however to which we have called attention above — emancipation from primitive ideas and absence of national feeling — the ex- planation is to be found in contact with civilised communities, especially in the form of mercenary service. The military life further had the effect of making the kings regard themselves primarily as commanders of armies. It was for their warriors that consideration was required rather than for the tillers of the soil, who were largely of alien nationality. When this is realised it becomes easy to understand the instability of heroic society. The military followers of a peace- loving king, unless he was very wealthy and generous, were liable to drift away, while the bulk of the population counted for nothing. In the absence of any truly national organisation or national feeling all depended on the personal qualities of the leaders. Under Theodric the Ostrogoths were the chief power in Europe ; but within thirty years of his death they disappear and are not heard of again. Under Dusan the Servians seemed destined to absorb all that was left of the Greek empire ; after his death they failed to offer any effective resistance to the ^ We may compare the force led by Agamemnon against Troy and the great mixed host encountered by Rameses IH (of. p. 188). 462 THE CAUSES OF THE HEROIC AGE [CHAP. Turks. The kingdoms of the Greek Heroic Age seem to have succumbed to much less formidable antagonists. So numerous indeed are cases of this kind that one is perhaps justified in regarding national disaster as the normal ending of such epochs. The Heroic Age, both Greek and Teutonic, presents us with the picture of a society largely free from restraint of any kind. In the higher ranks tribal law has ceased to maintain its force; and its decay leaves the individual free from obligations both to the kindred and to the community. He may disregard the bonds of kinship even to the extent of taking a kinsman's life ; and he recognises no authority beyond that of the lord whose service he has entered. The same freedom is exhibited in his attitude to the deities. It is of course in princes that we find these features most strongly developed. That which they prize above all else is the ability to indulge their desires to the full — in feasting and every form of enjoyment for themselves, in unlimited generosity to their friends, in ferocious vindictiveness towards their foes. The hero of the Odyssey, when his opportunity arrives, sets no limit to the vengeance which he exacts, from prince, goatherd and maidservant. Achilles, the chief hero of the Iliad, is trans- formed into a savage when he gets possession of the dead body of his enemy. His story furnishes a fitting parallel to that of Alboin, whosfe brutal conduct brought upon him so swift a retribution. And it is to be remembered that this Alboin's generosity was a theme of poetry from Italy to England. The best side of heroic kingship may be seen in such a character as Hrothgar. His conception- of the duties of a national ruler may have been of a somewhat elementary character. But it is rather as the head of a large household that we have to regard him ; and as such he commands our esteem. Even in the Merovingian family — we may cite Gregory's description (HI 25) of Theodberht — there were princes who won the respect of Roman ecclesiastics. In the courts of such princes the conditions of life were probably as good as at any time for many centuries later. We have no reason for sup- posing that the case was otherwise in the Heroic Age of Greece. XIX] THE CAUSES OF THE HEROIC AGE 463 But above all we have to remember the heroic poems. It is not reasonable to regard the Anglo-Saxon poems, much less the Homeric poems, as products of barbarism. The courts which gave birth to such poetry must have appropriated to a considerable extent the culture, as well as the wealth and luxury, of earlier civilisations. It is to be remarked however that the hold which these poems have exercised on subsequent ages, in very different stages of culture, is due not only to their artistic qualities but also to the absorbing interest of the situations which they depict. This interest arises very largely from the extraordinary freedom from restraint enjoyed by the characters in the gratification of their feelings and desires and from the tremendous and sudden vicissitudes of fortune to which they are exposed. The pictures presented to us are those of persons by no means ignorant of the pleasures and even the refinements of civilised life, yet dominated by the pride and passions which spring from an entirely reckless individualism and untrained by experience to exercise modera- tion. According to the view put forward above the explanation of such features is to be found not so much in any peculiarly fertile gift of imagination by which the conventional court poetry of these periods was inspired, but rather in the circum- stances of the times and in the character of the courts which produced that poetry. ADDENDA ET CORRIGENDA. Page 19 ff. For further information the reader may be referred to Miss Clarke's Sidelights on Teutonic History during the Migration Period (Cambridge, 191 1), which contains a very clear and interesting account of the various characters mentioned in the heroic poems. Page 43, 11. 5 — 7. This suggestion can hardly be maintained. The true name of Theodberht's son was probably Theodwald. Page 46, 1. 3 ff. The consideration of this difficult question has recently been somewhat facilitated by Richter's Chronologische Studien zur ags. Literatur (Halle, 19 10). The general effect of Dr Richter's investigations is to confirm the view put forward by Prof. Sarazzin {Engl. Stud., xvill 170 ff.) as to the antiquity of Genesis A. Unfortunately I fear that the statistics are not complete and, further, that the evidence is not always treated with strict impartiality. Thus in Beowulf such half-verses as to widan feore are regarded as proofs of shortening (through loss of h) and reckoned in the final statistics (pp. 9, 85) ; but in Genesis A the metrically equivalent on fyore lifde is not so reckoned (pp. 24, 89). The half-verse geseon meahton is cited in Beowulf without qualification (p. 15), but in Exodus as doubtful (p. 18). In Beowulf -wundor seon is taken to be - x - (pp. 13, 15), but in Daniel A it is treated as doubtful (p. 32). Dr Richter's conclusion that Genesis A is an earlier poem than Beowulf is certainly not substantiated by the treatment ol feore or of postconsonantal r, /, m, n^, or again by that of frea, don, gan, sie^ ; while compounds such as \reanyd obviously do not stand on the same footing as case-forms like \rea. It is only in the treatment of intervocalic h that Genesis A apparently shows a more archaic character than Beowulf In the former poem Dr Richter cites only one case of contraction (p. 28), and even this is doubtful ; but the same remark applies to at least 18 of the 24 (genuine) cases which he cites for Beowulf (p. 15). Out of the five or six probable cases of contraction in this poem three occur in practically the same phrase — in {on, to) sele pam hean — while two of the others occur in consecutive verses (910 f). The conclusion to which the evidence seems to me to point is that both Genesis A and Beowulf (even in its Christianised form) date from the seventh century, but that the former has been somewhat better preserved than the latter. As Genesis A is doubtless of monastic origin, we may reasonably expect that it was committed to writing at an earlier date. In the intervening period the text of Beowulf may have suffered many changes (such as the insertion of the article) at the hands of minstrels. ^' ^ As shown by Dr Richter's lists (pp. 9 ff., 24 ff. and 13 ff., 27 f.). ADDENDA ET CORRIGENDA 465 The further question raised by Prof. Sarazzin {op. cit., p. 192 ff.) as to the relationship of Genesis A to Caedmon cannot be discussed here. But in view of the evidence brought forward I do think it would be worth while to examine and compare the characteristics of the various ' Caedmonic ' poems from all points of view. Apart from the extremely improbable dating of certain sound-changes proposed by Prof. Morsbach (cf p. 66 ff. above) and accepted both by Sarazzin and Richter, I see no reason for supposing that Caedmon's poems have entirely perished. Page 61, 1. 4 ff. Throughout this book I have followed the view generally held by scholars in the past that the expression geongum cempan in Beow. 2044 refers to Ingeld. In recent years several scholars have adopted a different interpretation, viz. that it means not the prince himself but a nameless member of his retinue. This interpretation is due primarily to V. 2061 f., where it has been assumed, somewhat hastily in my opinion, (i) that se <7Si?r must denote the slayer and (2) that Ingeld's own country is the scene of the tragedy. On the basis of this interpretation Prof. Olrik {Danmarks Heltedigtning.^ ll p. 30 ff., especially p. 37 ff.) has now put forward a theory that the episode in Beowulf has little more than the names in common with Saxo's story of Ingellus' revenge, and that the latter is in its main features the creation of a Danish poet of Harold Bluetooth's time. His explanation is certainly interesting and ingenious ; but at the same time it appears to me to be open to serious objections. Thus he has to admit (p. 39) that in the only other reference to the bridal tragedy ('bryllups- kampen') contained in the poem (v. 82 ff.) the scene is laid in the Danish king's hall. His explanation is that here we have a variant form of the story (cf. Widsith, v. 45 ff.), emanating from a different poet. But surely one cannot place much confidence in an interpretation which involves the assumption of such an inconsistency as this^ Still more important is the reference to Ingeld {Htnieldus) in Alcuin's letter to Hygebald (cf. p. 41), which Prof. Olrik seems to have entirely ignored. From this passage it would seem that Ingeld's fame was as great in England in the eighth century as it ever became in Denmark ; and such fame is scarcely compatible with the passive role assigned to him by the theory under discussion. But, more than this, have we any ground for supposing that Alcuin did not exercise his judgment in taking Ingeld as his example of a perditus rexl If the story of Ingeld known to him was substantially identical with that ^ So far as I can see the only argument for this inconsistency which Prof. Olrik brings forward is the statement (Vol. i, p. 16; cf. Vol. II, p. 38, note i, and p. 39, note i) that the fight at the marriage precedes Beowulfs visit to the king's hall. This however seems to mean that the present tense, which is used throughout the episode (nearly a score of examples), must be taken as a historic present — a construction which is rarely or never found elsewhere in Anglo-Saxon heroic poetry. In Beowulf only one instance (v. 1879) is cited by Nader (Anglia, X 547), and this is clearly erroneous. A possible case does occur in v. 1923 (wurta^); but most recent editors either emend (to wunade) or regard the passage as a speech. 466 ADDENDA ET CORRIGENDA recorded by Saxo, as I myself believe, he could hardly have found a case in which Christian and heathen obligations presented a more glaring contrast. To the heathen this hero appealed as the pious son who exacts vengeance for his father's death ; but to the Christian he was doubly abhorrent, not merely as a heathen but also as a murderer and a truce-breaker. Like other heroic stories recorded by Saxo and Icelandic authors the story of Ingellus has without doubt a long poetic history behind it. Prof. Olrik's investigations have rendered it probable that an element in this poetry dates from the tenth century. It may be that the poem used by Saxo was actually composed — or re-cast — at that time. But I must confess to some scepticism as to the possibility of determining with certainty what proportion of the material is to be assigned to the various periods inter- mediate between the sixth and twelfth centuries ; and I am disposed to think that in general Prof. Olrik has underrated the significance of the earlier elements. Page 64, I. 6. It should have been mentioned that the expression "to persons born after 800" applies only to the period extending down to 975. I have not examined the evidence available for later times. About this time a new factor is introduced by the revival of interest in heroic poetry. I would also call attention to the fact that the statistics are confined to historical documents and do not include names (of moneyers) found only on coins. My impression from a perusal of the material collected in Mr Searle's Onotnasiicon Anglo- Saxonicum is that this element will not appreciably affect the statistics. Page 75, I. 26fTf. Reference should have been given to the Exeter Gnomic Verses, Sgfif. Page no. On the subject discussed at the beginning of this chapter the reader is referred to Heusler, S.-B. d. Akad. zu Berlin^ 1909, P- 937 fi"-, and Van Gennep, La Formation des Lcgendes. The definitions given above may not be entirely satisfactory — mention should perhaps have been made of the ' Ortssage ''■ — but they will probably be found sufficient for the purpose of this book. Page ii6ff. On this subject reference should be made to Panzer, Studien zur germ. Sagengeschichte, I Beowulf ., which contains a careful and detailed examination of the Marchen in question. Unfortunately I have not been able to use this work, as it did not come into my hands until the first eight chapters of my book were in print. In regard to those points on which I have chiefly laid stress, viz. that Beowulf is to be identified with Biarki and that he appears to be a historical character, Prof. Panzer's views (pp. 368 ff., 390 ff.) are practically identical with those expressed above. On the other hand he holds that the story of Biarki — the portion dealing with his origin, as well as the adventure at Leire — comes from the same folk-tale as the story of Beowulf and Grendel. His theory as a whole gives rise to important questions regarding the origin and distribution of folk-tales, which ADDENDA ET CORRIGENDA 467 cannot be discussed here. Thus I should like to know how far the same folk-tale can arise independently in different regions. Again, the Marchen with which he deals seem to me to be highly composite structures, which I should be inclined to regard rather as aggregations of folk-tales. This is true more especially of the elements common to the stories of Beowulf, Grettir and Ormr, the close affinities of which have been brought into a much clearer light by Prof. Panzer's researches. It may be that the elaborate story which underlies all three was treated in poetry at an early date ; but I see no more reason now than before for believing that the two Icelandic stories have been affected by any poem dealing with Beowulf himself Page iiyf. It is perhaps worth noting that Ormr Storolfsson was related to Grettir, both being descended from Haengr, the great-grandson of that Ketill Haengr of Hrafnista whose story is largely taken up with adventures with trolls and iotnar. Page 124. For the death of Frotho III the reader may be referred to Olrik, Danmarks Heltedigtning, ll p. 239 ff. ; and for the connection between this character and the Beaw of the genealogies to {ib.) p. 249 ff. As stated above (p. 126, note) I should prefer the form Beowa, for earlier Biowi {diS Aella for earlier Aelli; cf p. 64). The latter form may be con- nected with Byggvir. Page 139 ff. On the Nibelungenlied reference may be made to Prof. Rothe's article Nibelungias und Waltharms (S.-B. d. Akad. zu Berlin, 1909? p. 649). Page 161, 1. I f The references to Sarus (accidentally omitted here} are Olympiodoros, p. 449 (ed. Niebuhr), Jordanes, Roinana, § 321 {Mon. Germ., Auct. Ant. v 41). Page 180, 1. 20 ff. It has rightly been pointed out to me by Mr C. A. Scutt that, if the statement attributed by Herodotus (Vi 3) to Demaratos is to be trusted, the average length of a generation in the Spartan royal families may reasonably be expected to be somewhat longer than elsewhere. Yet in point of fact this is not borne out by the evidence available for the historical period (from the fifth to the third centuries), nor yet by that of the genealogies for the period between Theopompos and Cleomenes, if the commonly accepted date for the first Messenian war is correct. Page 185, 1. 20 ff. Recent discoveries have rendered it probable that cremation was practised in Crete in very early times. Up to now however all the evidence apparently comes from the Early Minoan period, so that the point upon which Dr Mackenzie lays stress remains practically un- affected. Page 190, 1. 2ff. I regret that when this was written I had not observed that Prof. Meyer accepts the statement of Xanthos with regard to the Phrygian invasion. On this question see p. 437, note 2. 468 ADDENDA ET CORRIGENDA Page 198, 1. 27 ff. On this subject reference may be made to Shewan, The Lay of Dolon^ which appeared too late for me to be able to make use of it. Page 244, note 2. A different view is taken by Allen, Jour. Hell. St. XXX 312 ff., where a full discussion of the Trojan Catalogue — as also of the Achaean Catalogue {ib.., p. 292 ff.) — will be found. Mr Allen's conclusions differ greatly from the views expressed above. Page 265, note 3. In the poem HyndlulioS (cf. p. 12) the genealogy of Ottarr the son of Innsteinn is traced back to a certain Svanr hinn rauSi. The same genealogy occurs in the document Hversu Noregr byg^ist (published in the Fornaldar Sogur NorSrlanda, II p. 6 in Dr Valdimar Asmundarson's edition) ; and here Svanr hinn rauSi is said to be the son of Finndlfr by Svanhildr the daughter of Day {Dagr Dellingsson) and the Sun {Sol, dottir Mundilfara). This document however belongs to a very late period — the close of the fourteenth century — and I know of no earlier authority for the first part of the genealogy. Page 285, note i, 1. 4. For "l p. 687" read "i ii p. 687," and similarly in all subsequent references to the second edition of Prof. Meyer's Geschichte des AUe?-tums. Page 313, note i. For "Pasic" read "Pavic" (and so also in the notes on the following pages). INDEX Achaia 174, 281, 289. Achaioi (Achaeans) 176 f., etc., (in Crete) 177 f., 186 f. Achilles 175, -112, 250, 263 f., 269, 275 ff., 288 f., 295 f. , 462. A'Sils (cf. Eadgils) 17, 20, 31, 153 f. Aelfwine (see Alboin). Aeolic element in Homeric poems 208, 215' 273, 290 f. Aeolic peoples, etc. 209, 272 ff., 295 f. Aeolis 210 f., 227, 231 ff., 240. Aeschylus 171, 237 ff Agamemnon (i) of Cyme 227, 231. Agamemnon (ii) of Mycenae 174, 222, 237 ff., 274 ff., 288 f., 302, 382 ff. Aias (i), son of Oileus 175, 302. Aias (ii), son of Telamon 175, 202, 204, 302 f. Aithiopis 170, 197, 236, 244. Alboin 26, 28, 462. Alcinoos 384. Alcman 215, 222, 266. Alcuin 41, 73, 79 f., 465. Aldhelm 80. Ali (cf. Onela) 20, 153 f. alliances 373 f., 387 ff. Amalaswintha 371 f. Ammianus Marcellinus 23, 366. Amyclai 240, 302. Angli 31 {., 51, 61, 97 f., 376, 405. Ansgar 255 f. Aphrodite 416 ff. Apollo 250 f., 416 ff. Arcadian 284 f., 287, 436 f., 453. Archilochos 194, 216, 226 f. Ares 416 ff. Argo 196, 245, 261. Argos 274 ff. Artemis 416. Asgar*Sr 400. Asia Minor (ethnography of) 178, 245 ff. assembly 368 f., 382 f. Athene 250 f., 416 ff. Athens 171, 216 ff., 381, 421. Atli (cf. Attila) 13 f., 22, 146 f. Attila 22, 37 f., 84 f., 156. Audoin 371. Balder 112, 400 f., 405. Bavarian characters 31, 33. Bede 71, 73, 79, 350, 414 f. Bellerophon 212, 356 f., 380. Beowulf I f., 47 ff., etc. Bemlef 80, 87, 222, 243. Biarkamal 17, 60, 121, 402. Biarki 119 ff., 466. birds (speaking) 98, 128. Bithynians 245 f. boasting 326 f. Boeotia 214, 283 f. Bosnian poetry loi ff., 216, 429. Bragi 15, 58. Bravalla (Bravik) 25, 251. bronze and iron 195, 199 f. Brunanburh (poem) 3, 34, 332. Brynhildr 7, 13 f., 14O ff., 163 ff. Burgundians 7, 22, 31, 351, 366. Cadmos 267. Caedmon 45 f., 79, 230, 465. Caesar 413, 427 ff. Callinos 171, 226 f. Caria, Carians 179, 245 ff. Cassiodorus 26, 85, 374. Catalogue (Hesiodic) 194, 197 f., 227, 242. Catalogue of Ships 174, 279, 289, 379, 387 f., 468.^ Chalon-sur-Saone 162. chariot 203 f. Charles M artel 127 f. Chios 176, 209. 470 INDEX Christian influence in Anglo-Saxon poems 47 ff., 241. Cilices 246 ff. Cinaithon 170. Circe 258 f., 419. class gradations 351 f., 364 f. Cleisthenes 193. Clovis 31, 85, 347. Clytaimnestra 229, 237 f. Cnossos 178, 311. comitatus 328 f., 348, 350, 361 ff., 376 f., 443- council 369 ff., 383 ff. cremation 53 ff., 397 f., 410 f., 422 f., 467. Crete 177 f., 184 ff., 311 ff., 435, 450 ff., 467. Cumbrian history 108 f., 430, 446 f. Cyme 214. Cynewulf (i), poet 45. Cynewulf (ii), king of Wessex 349, 351. Cynurioi 460. Cypria 170, 197, 234 ff., 244, 248. Cyzicos 196. Danes i f., 15, 17, 20 f., 24f., 29, 31, etc. Danish heroes i f., 17, 20, 31, etc. Demodocos 221, 234. Deor 3, 56, 81, 92. Dietrich von Bern (cf. Theodric) 5, 7 ff., 23. 154 f-. 3" f-. 333. 452- Diomedeia 250 f. Diomedes 175, 278, 379 f., 419. divine parentage 264, 357. divinity of kings 367, 381 f. Dodona 417, 437. Doloneia 198 f., 298. Dorians 174, 178 f., 184, 186 f., 459 dragons 122 if. Dusan 430, 448, 461. Eadgils (i) 2, 27, 32, 56, 81. Eadgils (ii) (cf. AiSils) 2, 20, 153. Eastgota (see Ostrogotha). Edda (Older) 10 ff., 99 f., 232. Edda (Prose) 15. Edda theology 394, 400, 404, 406 f. Edwin 334, 349, 369. Egill Skallagrlmsson 16, 226. Egypt 184 f., 187 ff., 266, 455. Einhard 5, 62, 243. Eireksmal 15, 257. Ekkehard 6, 57, 162. Eormenric (Ermenrich) 2 f., 8 ff., 23, 37. 157. 347- Epeiros 437, 460. Epigonoi 171. eponymous heroes 131 f., 267 ff., 301. Eratosthenes 179. Eric the Victorious 252 f. Etruscans 449 f. Etzel (cf. Atli) 7 f., 22, etc. Exodus (Anglo-Saxon) 45, 464. Eyvindr Skaldaspillir 15 f., 91, 254. family pride 327 f. festivals 369, 383. fictitious names 44, 159 ff., 299. Finn 2, 20, 26, 42, 333. folk-tales no ff., 258 ff., 264 ff., 466 f. Franks 31, 39, 337, 343, 345 f., 352, 368, 460. Frey 11, 15, 125, 253, 367, 394 f., 399 ff., 404. Freyia 1 1 f., 400, 402. Fria (cf. Frigg) 115, 408. Frigg 394, 403. 405- Frisians 2, 31, 80, 89 f., 370. Fro'Si 15, 27, 123 ff., 467. Gaulish heroic poetry 427 f. Gaulish history 427 ff., 449 f. Gautreks Saga 239, 385, 403. Geatas (Gotar) i f., 31 f., 35, 119 f., etc. Gelimer 86, 94. genealogies, Anglo-Saxon 42, 408 f. genealogies, Greek 179 ff., 267. Genesis (Anglo-Saxon) 21 r, 464. Genseric 369 f., 460. Geruthus 258 f. Glaucos 341. glory, love of 87 f., 325 f., 329. glossaries, Anglo-Saxon 68, 70. Gliimr 253. Gondul 254, 263. Gormo 258 ff. Gothic heroic poetry 10, 62 f., 86 f. Goths 23, 26, 31 ff., etc. Grendel 116 f., 121 f., 466. Grettis Saga n6 f., 467. Gu^run 13 f., 37, 146 f., 152 f., 156, 163 ff. Gundobad 22, 338 f. INDEX 471 Gunnarr (see Guthhere) Gunnlaugs Saga 91 f. Guntharius, Gunther (see Guthhere). Guthhere (Gunnarr, Gunther) 6 f., I4, 21 f., 138 f., 141 ff., 146 f., 163, 165. Gylfaginning 229. Haakon, earl of Lade 16, 91, 252 ff., 401 f. Haakon I, king of Norway 15, 91, 254, 263. Hades 422 f. Hadingus 125 f., 260. Haethcyn 337, 346. Hagbar«r 17, 27, 333. Hagena (Hogni) (i), 6 f., 14, 137 f., 140 ff., 161 f. Hagena (Hdgni) (ii), see He'Sinn and Hogni. Hakonarmil 15, 97, 254 f. Ham'Sir and Sdrli 14, 37, 156 f., 160 f., 333 (cf. Sarus and Ammius). Haraldr Hilditonn 17, 24 f., 251. Harold the Fair-haired 15, 91, 296. Havamal 11, 229, 397. heathen elements in Beowull 52 f. Hector 269 ff. He'Sinn and Hogni 16, 21, 27 f., 112, 135 ff-i 333 (cf. Hetel and Hagen). Helen 249, 265 ff., 337. Helios (cattle of) 259. Hellenic peoples 285 ff. Hephaistos 416 f. Hera 416 ff. Heracleidai, Return of the 174, 180, 331. Heremod 20, 60, 366. Herhor 312. Hermanaricus 10, 19, etc. (cf. Eormenric). Hermes 416. HeruU 346, 372 f., 407, 411, 445. Hesiod 171 f., 194, 210, 214 f., 223, 228 ff., 242. Hetel and Hagen (cf. He^inn and Hogni) 8f. , 21, etc. Hildr (Hilde) 8 f., 16, 136 f. Hildebrandslied 5, 62, 94 f. Hiorungavagr 252. Hittites 178, 188 ff., 204 f., 246 f. Holmryge (cf. Rugii) 21, 28. Homer's birthplace 179, 208 f. Homeric armature 185 f., 191, 201 ff. Homeric civilisation 186 f., 199 ff., 434 ff- Hornklofi 15, 91. Hrolfr Kraki 17, 20, 349. Hrolfs Saga Kraka 120 ff. Hrothgar i f., 20, 83, 119, 354, 368, 462. Hygd 372. Hygelac i f., 24, 341, 371 f. HyndluIiotS 12, 229, 402, 468. lason 196. Iceland 10, 91 ff., 394 f.i 398 ff. Idomeneus 175, 278, 306. Iliad 169, 174 f., 193 ff., etc. images 205. Ingeld (Ingellus, Ingialdr) 17, 36, 41 f., 61. 333. 465 f- inscriptions, Anglo-Saxon 69, 71, 214. inscriptions, Greek 209, 217, 280 f. international interest of heroic poetry 34 ff-, 335 f- international relations 373 ff., 386 ff. invulnerability 129, 264. Ionian Isles 175, 274. Ionic cities 176, 236, 245 f. Ionic dialects 208 f., 215 ff., 283. Ionic language of Homeric poems 208, 215 ff., 273. lormunrekr 14, 19, etc. (cf. Eormenric). Iphigeneia 238 f. Istros 269 f. Ithaca 356, 378. Jomsvikingar 251. Jordanes 10, 26, 62, 86 f., 157. kindred 344 ff., 353 ff., 442 f. kingship 227 f., 337, 366 ff., 378 ff. kinsman (slaying of) 346 ff., 359 f. Kossovo 313, 316 f., 429, 441. Kraljevic Marko 441. Kriemhild (cf. Gu'Srun) 7 f., 19, 145 ff., 163 ff- Kudrun 8 f., 16, 19, 34, 135 f. Langobardi 9, 26, 62, 89, 115, 371 ff. Langobardic poetry 9 f., 115, 251, 413. Langres 31, 162. Lazar 313 ff. Leire 377, 403. Leleges 246, 248. Lesbos 176. 472 INDEX Little Iliad 170, 197, 236. Ljubovic (Beg) 103, 441. Locrians 269 f., 278, 357. Lo'Sbrok (Ragnarr) 17, 24, 295 f. Lokasenna ii, 83, 403. Loki II, 403. Ludwigslied 5, 21, 34, Lycians 178 f., 212, 356 f. Lycurgos 334 f. Maldon (poem) 3, 97. manes-worship 398 f., 410, 421. Marcellinus Comes 37. marriage-customs 98, 205, 358 f. Menelaos 175, 277, 356, 362. Merenptah 187, 455. mercenary seivice 445 ff., 455 f., 461. Merovingians 345 f., 367. Merseburg charms 405, 408 f., 412. Midas 227. Miletos 176, 231, 246. Milos Obilic 314 f., 317 f- Mimnermos 332. Minoan civilisation 184 ff., 451 ft. Minos 310 ff^. minstrelsy 79 ff-, 221 ff. Murad I 313 ff. Mycenae 185, 240, 274, 308 f., 435, 450 ff. Mycenean civilisation 185 ff. , 190 ff., 200 ff., 218, 233, 435 ff., 451 f. Nekyia 197, 229, 257, 260 f. Nestor 175, 272, 277, 342. Nibelungenlied 7, 139 ff., 164 f., 224, 233, .393. 467- Niflungar (Nibelunge) 140 ff. Nior'Sr 399 f., 405. nobility 352, 364. north-west Greek 281 ff. Norway 10, 15, 30, 32 f., 59, 91 f., 100, ii8ff., 375, 395, 401, 404, 406 f. Nostoi 170, 197, 234 ff. Odoacer 23, 376 f. Odysseus 175, 258, 273 f., 288 f., 297 ff., 358, 363 f., 419. Odyssey 169, 175, 193 ff-, etc. Offa (i), king of Angel 4, 17, 26 f., 126, 333. 376. Offa (ii), king of Mercia 126, 334 f. officials 352. Ohthere (Ottarr) 20, 153. Old Saxons 88 f., 370, 407, 410. Olympos 232, 416 f. Onela (cf. AH) 2, 20, 153 f., 347. Ongentheo 2, 153, 340. Orestes 237. Orms Thattr Storolfssonar 117 f., 467. Ostrogotha 10, 21, 26. Othin II, 13, 114, 251, 253 ff., 394 ff. ^ (cf. Woden). Ottarr (i), son of Innsteinn 12, 402, 468. Ottarr (ii), father of A'Sils (see Ohthere). Paphlagonia 245. Paris 222, 229, 269 ff. Patroclos 328 f., 361 f. patronymics 355. Paulus Diaconus 25 f., 372 f. Peisistratos 216, 256. Pelasgic Argos 274, 278 ff. Pelasgoi 177, 186, 248, 285 ff. Penelope 358 f. Pentaur (Poem of) 247, 455. Periandros 261, 422. personal allegiance 348 ff. , 361 ff. personal names (heroic) 42 ff., 64 ff., 231. Phaeacians 297, 299 f., 384 f. Phaistos 268, 311. Phemios 221, 234. Philistines (see Pulesatha). Phocis 283. Phoinix 279, 360 f., 363. Phthia 279, 286. Phthiotis 176, 280, 282, 286 f. place-names (heroic) 43 f. Poseidon 416 ff. Polyphemos 258. Praisos 178, 213. priests 367, 380- Priscus 37, 62, 84. Procopius 86, 97 ff. , 128 ff., etc. Proitos 212, 452. Pulesatha 188 ff., 455 f. Pylos 277, 451. Quedlinburg Annals 6, 9, 23, 37, 80. Rabenschlacht 8, 23. Radiger -98, 370 f. Rameses II 188, 247, 455. INDEX 473 Rameses III 185, 188 f., 455. Ravenna 23, 452. Reginsmal, Fafnismdl, Sigrdrifumal 13, 145. 235. riding 204. Roman influence 444 ff. royal marriages 373 f., 386 f. Rugii 28, 31, 134 (cf. Holmryge). Russian heroic poetry 104 f. Salamis 193, 302 f. sanctuaries 205 f. , 400 f. , 406 f. Sarpedon 305, 357. Sarus (i) 160 f., 377, 467. Sarus (ii) and Ammius 10, 14, etc. (cf. HamSir and Sorli). Saxo Grammaticus 17, 20, 25, etc. Scheria 379, 384. Scyld 75, 131 f., 373. Servian (Christian) heroic poetry 103 f., 313 ff., 429 f. Servian history 313 ff., 430, 447 ff. Sety I 204. Seyfridslied 7, 122, 140 f., 144 f. Shardina 188 ff., 201, 451, 454 ff. shield 201 f. Shield of Heracles 169, 175. Sidonius ApoUinaris 85. Siegfried 7, 19, 31, 33, 139 ff., 164 (cf. Sigur'Sr). Sigarr 17, 21. Sigmundr (Sigemund) 2, 16, 59 f., 114, 122 ff., 142, 251, 297. Sigur^r 12 ff., 59, 114, 139 ff., 163 ff., 263 f., 333 (cf- Siegfried). Sigur'Sr Hringr 17, 24 f., 251. single combats 339 f. Smyrna 208 ff., 332. Sparta 179 ff., 184, 331. Spercheios 206, 279 ff., 287, 417, 419. spoils 340 f. stages in history of heroic poetry 94 ff. Starka^r 17, 36, 92 f. Stesichoros 171, 239 f., 266, 310. Strassburg, battle of 340, 349. succession 356 suitors of Penelope 300 f., 363. Sunilda 10, 160 f. (cf. Svanhildr). supernatural beings 114 ff., 250 ff. supernatural properties 128 f., 262. Svanhildr 14, 16, 37, 157 (cf. Sunilda). Swedes (Sweon, Sviar) 2, 15, 20, 31, 153 f-. 367 ff-. 399- Tacitus 78, 349, 368 f., 413. Tchakaray (Zakar) 188 f. Telemachos 242 f. Tell-el-Amarna 188, 192, 455. Thebais 171. Theodberht 23 f., 462. Theodric 3, 19, 23, 25, 29 ff., 154 f., 373, 461 (cf. Dietrich von Bern). Thersites 228. Theseus 266. Thessaly 232, 269 ff. Thetis 264, 418 f. Thi'Sreks Saga af Bern 9, 57, 134, 138 f., 145 f- Thiotiolfr 15, 91. Thor II, 253, 394 ff., 404. Thorger'Sr Holgabru'Sr 252, 401 f. Thorkillus 258 f. Thraco-Phrygian peoples 178 f., 190, 457 f- Thucydides 286, 293, 308 ff. Tiryns 452, 454. tragic poets 171 f. tribal heroes 267 ff. Troy (Hissarlik) 200, 294 f., 309. Tyrtaios 226, 331. Uffo 17, 61. Unferth 159, 347. Upsala 154, 206, 368, 377, 399 f.. 407. Valhalla (Valholl) 254, 396 ff, 410 ff Valkyries 254, 397, 411 f. Vandals 31, 86, 115, 369, 371, 445, 460. Vellekla 16, 254. Venantius Fortunatus 85. vengeance 344 f., 353. Vidigoia 27, 62 (cf. Widia). Viking Age 25, 90 ff., 100, 251 ff., 342, 404, 406 f. Virgil 74 ff. Volsunga Saga 16, 145 ff., 164, 251. Volundarkvi^a 12, 132, 134 f- Voluspa II, 229. Waldhere (Waltharius) 2, 6, 38 f., 57 f., 137 ff., 151 f., 161 f., 333, 341, 376 f. 474 INDEX Waltharius (see Waldhere). Wanderer (poem) 3, 92, 350. Warni 97 f. Warrior Vase 185, 191, 451 f. Weland 3, 9, 12, 132 ff. Welsh heroic poetry 105 ff. Weoxtan 20, 159, 3-29 f., 340, 347, 350- wergelds 345, 352, 354, 365. Widia 27, 132, 135 (cf. Vidigoia, Witege, Wudga). Widsith 2, 20 f., 26, 56, 81, 87 f. Widsith (name) 43 f. Wiglaf 159, 328 f., 349. Willebrord 259. Witege 8 f., 19, 27, 31, 132, 135 (cf. Widia). Woden (Wodan) 115, 405, 408 ff. (cf. Othin). Wolfdietrich 8, 23, 155. women in the Heroic Age 337 f., 372. Worms 6 f., 146. writing 211 ff. Wudga 19 f., 27, 331 (cf. Widia). Xanten 7, 31. Xanthos 437 f., 467. Ynglinga Saga 16, 367, 397, 399. Yngvi 367, 405. Zakar (see Tchakaray). Zeus 416 ff. Zeus Agamemnon 381. DATE DUE EH Correnf occrui -T item. #LO-45102 "^^/pvv^ ass: f^e> 2i\^ itJC SOUTHERN REGIONAL LIBRARY FACILITY AA 001 245 794 3 1210 00207 1932 V,