^{page 66 note 1} On this name and the various forms in which it occurs, see Napier and Stevenson, Crawford Collection, p. 143 ; O. von Feilitzen, The pre-conquest personal names of Domesday Book, p. 299; D. Whitelock in the Viking Society's Saga-Book, xii. 133–4.

^{page 66 note 2} By 974 at the latest, see Bjarni Aðalbjarnarson's ed. of Heimskringla (i, p. xcii).

^{page 66 note 3} It cannot be too clearly emphasised that the verses of the skalds, who composed for the kings of Norway and Denmark in the tenth and eleventh centuries, and occasionally for prominent noblemen, are preserved only in quotations in the Old Norse Sagas, particularly in Heimskringla and Fagrshinna. Hence, although the basic principle of the study of early Scandinavian history must always be to study the verses separately from the prose in which they are embedded, and to see if they necessarily bear the meaning which the prose alleges them to do, yet, even when this is done, the danger always remains that a verse may not be genuinely early, or may be early but not refer to the events with which the prose connects it. The verses in the Sagas of the kings seem to be given in good faith by the compilers, who appear to avoid the practice, which is not uncommon in other Sagas, of writing verses to fit their narrative, and alleging that characters in their story composed them.

^{page 66 note 4} Edited in Skjaldedigtning, IA, pp. 200 ff. ; IB, pp. 190 ff.

^{page 66 note 5} Ó1áfs Saga Tryggvasonar, chaps. 20, 89–90.

^{page 66 note 6} Pp. 105, 136–7.

^{page 67 note 1} Skjaldedigtning, IA, pp. 202 ff. ; IB, pp. 193 ff.

^{page 67 note 2} The best discussion of this lost Saga is that of Bjarni Aðalbjarnarson Om de norske kongers sagaer, pp. 199–201, 217–24; cf. van Eeden, W., Neophilologus, xxxi. 76–8Google Scholar, for a different view.

^{page 67 note 3} In the uncertainty, which now prevails, concerning the precise political conditions in Scandinavia in the time of Earl Hákon, it is not possible to estimate the likelihood of this story, that Eiríkr received a grant of Norwegian territory from the Danish king. It is not necessary so to interpret the verse quoted by Heimskringla to support the story : see Bjarni Aðalbjarnarson, Heimskringla, i. 250–1.

^{page 67 note 4} This is to be dated in all probability some time between 980 and 990: see Bjarni Aðalbjarnarson, op. cit., pp. cix–xii, where the question of the identity of the invaders is also briefly discussed. The latter problem does not concern the present enquiry.

^{page 67 note 5} Though he does not appear in Saxo Grammaticus, who also has an account of the battle (ed. Holder, p. 327). The accounts of the battle in Heimskringla, Fagrskinna and the various extant forms of Jómsvíkinga Saga no doubt represent combinations of material from Hlaðajarla Saga and the lost original form of Jómsvíkinga Saga, but it is now impossible to separate the elements. Eiríkr's presence in the battle is also mentioned in two thirteenth-century Norse poems about i t : one is referred to below, p. 73, note 4 ; the other is the Búadrápa of Thorkell Gíslason, edited Skjaldedigtning, IA, pp. 553 ff. ; IB, pp. 536 ff.

^{page 67 note 6} Tindr was an Icelander, who is stated by Fagrskinna and the A.M. 510 version of Jómsvíkinga Saga to have been present at the battle. These sources are not independent, however, and the statement may be a mere inference from the fact that Tindr described the battle. His poem is edited Skjaldedigtning, IA, pp. 144 ff.; IB, pp. 136 ff.

^{page 67 note 7} The poems of Thórthr are edited Skjaldedigtning, IA, pp. 212 ff. ; IB, pp. 202 ff. He was an Icelander, who was in Norway on various occasions, but the view that he visited England is pure supposition, based on the fact that he described Eiríkr's English campaign, though it is stated as a fact that he did so by the editors of the Crawford Collection (p. 145). He probably based his verses on Eiríkr's voyage to England and his campaign there on travellers' tales, and they are to be used with caution, cf. below, pp. 69–70.

^{page 68 note 1} In the earliest version of the Saga of Óláfr Tryggvason, written in Latin late in the twelfth century by Oddr Snorrason, an Icelander, and known to us from various recensions of an Old Norse translation (all edited by Jónsson, Finnur in Saga Óláfs Tryggvasonar, Copenhagen, 1932)Google Scholar, Eiríkr and his brother Sveinn chance to be absent from Norway, when Óláfr Tryggvason returns. The versions of the Saga in Fagrskinna and Heimskringla (both revisions of Oddr's Saga) make Eiríkr withdraw because of Óláfr's arrival, in order to conform with Thórthr's verses which they quote : Heimskringla assumes that Sveinn was with him, but Fagrskinna does not mention him. Of the Norwegian compendia of history, Ágrip makes the two brothers flee to Sweden on Óláfr's arrival, the Historia Norvegiae to Denmark, while Theodricus does not mention the matter. The first of these texts is edited by Finnur Jónsson, Halle, 1929; the other two in Storm's, G.Monumenia Historica Norvegiae, Christiania, 1880Google Scholar. They all belong to the late twelfth century, though Hist. Nor. is extant only in a later modified and extended form (thirteenth century). Ágrip makes use of both the others, and all three are influenced by Icelandic verse and tradition, so they cannot be regarded with confidence, when they agree with Icelandic sources, as giving confirmation of these from independent Norwegian tradition. The use of literary sources in these works must also be allowed for: e.g., Theodricus has material from William of Jumièges, and Hist. Nor. from Adam of Bremen; cf. also above, p. 56, note 5, and below, p. 78, note 1.

^{page 68 note 2} Eiríkr is curiously absent from the accounts of the fall of Oláfr Tryggvason in Adam of Bremen and Saxo Grammaticus, who do not give any information as to what arrangements Sveinn of Denmark made for the government of Norway.

^{page 68 note 3} Eiríkr's marriage is mentioned by Fagrskinna (p. 136) and Theodricus (Storm, p. 24). Heimskringla places it before the conquest of Norway in 1000 (Óláfs Saga Tryggvasonar, chap. 90). Heimskringla is the only source for Sveinn's marriage, which it places at the time of the conquest (ibid.:, chap. 113). These marriages were no doubt mentioned in Hlaðajarla Saga and hence found their way into Heimskringla and Fagrskinna. Theodricus may have known of Eiríkr's marriage from independent tradition. When Ágrip (ed. Finnur Jónsson, p. 24) and the Legendary Saga of Óláfr Helgi, chap. 10, call Eiríkr a relative of Knútr, they doubtless allude to his Danish marriage.

^{page 68 note 4} Also a good many other verses by Hallfrethr and Skúli Thorsteinsson, and the ninth verse of Eyjólfr's poem referred to above.

^{page 69 note 1} This was the case in the chronological system of Sæmundr Fróthi, which is known from the late twelfth-century Icelandic poem, Nóregs konunga-tal (see Skjaldedigtning, IA, pp. 579 ff.; IB, pp. 575 ff.), and used by Fagrskinna (p. 144) ; since Heimskringla also allows Eiríkr a rule of twelve years, we may presume that the system of Ari Fróthi here agreed with that of Sæmundr. Ágrip has the same system, and Theodricus, while allowing Eiríkr fifteen years, also has the gap of two years thereafter (Storm, p. 25).

^{page 69 note 2} So Nóregs konunga-tal, Theodricus, Ágrip, Heimskringla, Fagrskinna. Worthless additions to the story in the late expanded version of the Saga of Óláfr Tryggvason allege treachery to the part of Knútr or of an old enemy (Fornmanna Sögur, iii. 31 ; Flateyjarbók, i. 561). Theodricus differs from the other sources in regarding Eiríkr's departure from Norway as due to uneasy relationships with his brother, and here we may have a separate Norwegian tradition. Nóregs konunga-tal does not state why Eiríkr left Norway.

^{page 69 note 3} Óláfs Saga Helga, chap. 24.

^{page 69 note 4} Finnur Jónsson's translation of this verse in Skjaldedigtning is an absurdity. The sense is well given by Vigfússon and Powell (Corpus Poeticum Boreale, ii. 104).

^{page 69 note 5} As the Encomiast tells.

^{page 69 note 6} See above, note 2. Theodricus does not make himself clear, whether Knútr or Sveinn was the Danish commander in England when Eiríkr arrived.

^{page 69 note 7} Ed. by Petersens, Af and Olsen, , in Sögur Danakonunga (Copenhagen, 1919–25), pp. 44–5Google Scholar.

^{page 69 note 8} Óláfr certainly returned in the autumn of 1014 or 1015 (see below, p. 79, note 8) and the evidence points fairly strongly to the former year. If, however, it be assumed that Óláfr returned in 1015, it follows that Eiríkr may have joined Knútr in England, not in Denmark, and that Thórthr's account may be correct.

^{page 69 note 9} Old English Chronicle.

^{page 69 note 10} According to the version of the Saga of Óláfr Helgi known as the Legendary Saga, rumours of Óláfr's movements were already current before Eiríkr left Norway. The Legendary Saga, either independently or following the Oldest Saga, here inserts a piece of tradition which clashes with its main narrative, in which Knútr and Eiríkr are already in England before Óláfr leaves that country (see below, p. 80). See the note in Keyser and Unger's edition, p. 104. The Legendary Saga also makes Hákon pay a flying visit to England to ask Knútr's help as Óláfr approaches. On the nature of the Legendary Saga, and the relationships of the various Sagas of Óláfr Helgi, see below, pp. 80–1.

^{page 70 note 1} Old English Chronicle.

^{page 70 note 2} See N.C., i. 660.

^{page 70 note 3} Old English Chronicle.

^{page 70 note 4} N.C., loc. cit.

^{page 70 note 5} So Freeman suggests, N.C., loc. cit.

^{page 70 note 6} Viking Society's Saga Book, xii. 132.

^{page 70 note 7} Gesta Regum, ii. 181. Henry of Huntingdon follows William.

^{page 70 note 8} See references above, p. 69, note 2. Some of these sources say that Eiríkr was then going on a pilgrimage to Rome or had just returned. Heimskringla and Fagrskinna err in placing his death respectively one and two years after he came to England.

^{page 70 note 9} Óláfs Saga Helga, chap. 25.

^{page 70 note 10} See below, p. 77.

^{page 71 note 1} Eiríkr's presence at London during the siege and his being in some way related to Knútr (see above, p. 68, note 3) are the only facts concerning him, which clearly belong to the Saga of Óláfr Helgi in its early form, and are accordingly almost the only ones which appears in the Legendary Saga. (The only exception is the curious tradition recorded by the Legendary Saga which is discussed above, p. 69, note 10.) The additional information concerning his career after 1000, which we find in Fagrskinna and Heimskringla, comes from Hlaðajarla Saga, and Thórthr's verses.

^{page 71 note 2} See below,, pp. 76–7.

^{page 71 note 3} Steenstrup, Normannerne, iii. 284, n. 2, hints that he does not consider the verse genuine. The verse is palpably influenced by that of Óttarr, in which the battle of Ringmere of the year 1010 is referred to (see below, p. 77) : both verses have in common the line rauð Hringmaraheiði. Nevertheless, Finnur Jónsson (Knytlingasaga, dens Kilder og historiske Værd, Copenhagen, 1907, p. 17), prefers to regard the verse as genuine and compares the sporadic raiding attributed by the Encomiast (II, 7) to Eiríkr.

^{page 71 note 4} Storm, p. 24; Saga Óláfs Tryggvasonar, ed. Finnur Jónsson, pp. 220 ff.

^{page 71 note 5} See Adam of Bremen, ii. 39. Theodricus states that Eiríkr allowed freedom of religion, but the Historia Norvegiae that he and his brother nearly uprooted the faith (Storm, pp.,25 and 119). Heimskringla and Fagrskinna support Theodricus, and their agreement, as usual in matters concerning Eiríkr, points to Hlaðajarla Saga as their source.

^{page 72 note 1} See p. 85.

^{page 72 note 2} Theodricus (Storm, p. 31) ; Ágrip (ed. Finnur Jónsson, p. 29) ; and all forms of Óláfs Saga Helga.

^{page 72 note 3} The poem is known as Tøgdrápa : parts of it are quoted with reference to the events of 1028, by all the Sagas of Óláfr Helgi. Edited Skjaldedigtning, IA, pp. 322 ff. ; IB, pp. 298–9.

^{page 72 note 4} See Legendary Saga, chap. 76.

^{page 72 note 5} Theodricus (Storm, p. 31) ; Ágrip (ed. Finnur Jónsson, p. 29).

^{page 72 note 6} Óláfs Saga Helga, chaps. 146, 161.

^{page 72 note 7} P. 170.

^{page 72 note 8} Legendary Saga, chap. 77; Heimskringla's version, chap. 184; Fagrskinna, p. 179.

^{page 72 note 9} Theodricus (Storm, p. 34) ; Ágrip (ed. Finnur Jónsson, p. 29).

^{page 72 note 10} Chaps. 77 and 101 contradict one another. This is merely due to the fact that the Legendary Saga is interpolated from Ágrip. See Nordal, Om Olaf den helliges saga, pp. 34–5.

^{page 72 note 11} On his signature to K. 744, see above, p. 60.

^{page 73 note 1} Ágrip (ed. Finnur Jónsson, p. 26) has an unsupported and obviously impossible story that Óláfr Helgi made Hákon earl of the Sudreys, when he expelled him. from Norway.

^{page 73 note 2} Reference as above, p. 70, note 6.

^{page 73 note 3} That is not only in the five extant recensions, but in the accounts of the attempted invasion in Heimskringla and Fagrskinna which are derived from the lost early form of the Saga, from which the five extant forms are developed. How far these extant accounts draw also on Hlaðajarla Saga is a very obscure problem, see above, p. 67, note 5.

^{page 73 note 4} Died 1222. His poem is edited Skjaldedigtning, IIA, pp. 1 ff. ; IIB, pp. 1 ff.

^{page 73 note 5} Cf. above, p. 67, note 4, on the problem of the composition of the invading forces, a question which again does not effect this enquiry.

^{page 73 note 6} In a considerable number of verses : references in Lexicon poeticum antiquæ linguæ septentrionalis, Sveinbjörn Egilsson and Finnur Jónsson, s.v. Sigvaldi.

^{page 73 note 7} Ed. Holder, pp. 325 ff.

^{page 73 note 8} Edited Skjaldedigtning, IA, pp. 543 ff. ; IB, pp. 525 ff.

^{page 73 note 9} See Bjarni Aðalbjarnarson's edition of Heimskringla, i. pp. cxxxiii ff., for a brief consideration of this vexed question, and further references.

^{page 73 note 10} On the relationships of the sagas of Óláfr Helgi, see below, pp. 80–1.

^{page 74 note 1} This interesting passage is quoted N.C., i. 677–8 ; cf. above, p. lvi. Another sidelight on Thorkell's campaign is provided by Heremannus (Memorials of St. Edmund's Abbey, Rolls Series, i. 40).

^{page 74 note 2} We can safely reject William of Malmesbury's story (Gesta Regum, ii. 176) that Thorkell invited Sveinn to England in 1013, for we know that he loyally supported Æthelred during the invasion of that year. Cf. N.C., i. 668.

^{page 74 note 3} It is mentioned in the Old English Chronicle that Eadric Streona seduced forty ships to the Danish side late in 1015, and it is usually assumed that these were the remains of the forty-five ships of Thorkell's fleet, which entered English service in 1012. This is highly probable, for it seems unlikely that a native English fleet of such size was then in being. It is open to those who so wish to assume that Thorkell went over to Knútr with these ships.

^{page 74 note 4} N.C., i. 356.

^{page 74 note 5} Crawford Collétion, p. 141.

^{page 74 note 6} In § D of the present Appendix.

^{page 74 note 7} Pp. liv ff.

^{page 75 note 1} A further motive for the Encomiast's kindness to the memory of Thorkell is suggested below, p. 84, note 8.

^{page 75 note 2} On a statement in a worthless source that Thorkell was with Knútr in the invasion of 1015–16, see below, p. 88.

^{page 75 note 3} Also K. 742, dated 1026, but this is a ridiculous forgery. It is worth noting that a Đurkytel miles signs a charter of 1012 (K. 719 ; Codex Roffensis) ; this was the year in which Thorkell entered Æthelred's service, and it is very likely that this is his signature.

^{page 75 note 4} Liebermann, Gesetze, i. 273–5.

^{page 75 note 5} Thorpe, p. 308.

^{page 75 note 6} Reference above, p. 70, note 6.

^{page 75 note 7} Reference above, p. 55.

^{page 75 note 8} Old English Chronicle, MSS. C and D.

^{page 75 note 9} Ibid., 1021, and MS. C, 1023.

^{page 75 note 10} Old English Chronicle, MS. D ; cf. above, p. xlvii. The fact that Hörthaknútr was not physically committed to Thorkell's charge in 1023 has troubled various historians. Some have suggested that Haraldr was the son committed to Thorkell, others that the whole entry of Chronicle C for 1023 is a confusion, and that the incident referred to is the appointment of Úlfr as regent of Denmark and Hörthaknútr's guardian some years later (see below, p. 83). There is nothing to be said for either suggestion. It seems evident that there was some crisis in Knútr's dominions just after Thorkell was banished, for Knútr concentrated his fleet in 1022 at Wight, presumably to go in force to Denmark, where we find him in 1023. There is, however, no need to connect these events with Thorkell, or to regard his fall as more than a salutary lesson for a powerful subject, to be followed quickly by a restoration,. when its lesson had been learned.

^{page 76 note 1} Cf. below, p. 85. William of Malmesbury's statement (Gesta Regum, ii. 181) that Thorkell was murdered when he returned to Denmark, is derived from the worthless account of Thorkell given by Osbern (see N.C., i. 668–9).

^{page 76 note 2} See above, p. 75, note 3.

^{page 76 note 3} Ed. Thorpe, i. 183.

^{page 76 note 4} Her identity is discussed below, p. 89.

^{page 76 note 5} Pp. 84 ff.

^{page 76 note 6} Skjaldedigtning, IA, p. 220 ; IB, p. 210. Some of these curious snatches of verse, apparently by Norse soldiers who fought in England, are edited and discussed by Miss M. Ashdown, English and Norse Documents (Cambridge, 1930), pp. 140–3 and 205–8, though she does not include the one mentioning Thorkell. The general sense of it is given in Collingwood's translation, Scandinavian Britain, p. 157.

^{page 76 note 7} This poem and the similar one by Óttarr Svarti to be discussed below are edited Skjaldedigtning, IA, pp. 223 ff. and 290 ff. ; IB, pp. 213 ff. and 268 ff. The verses which concern Óláfr's English adventures may be conveniently consulted in Ashdown, op. cit., pp. 156 ff.

^{page 76 note 8} Ellu kind: see Ashdown, op. cit., p. 221.

^{page 77 note 1} Partar : see ibid., p. 222.

^{page 77 note 2} That Óláfr was active in Spain at this time seems fairly certain : see the two works referred to below, p. 79, note 8.

^{page 77 note 3} See the fragment of Styrmir's version of the saga of Óláfr Helgi in Flateyjarbók, iii. 242.

^{page 77 note 4} In Sigvatr's verse on the fighting at London there is an allusion to the defence of a ditch, and many historians have suggested that there is some confusion with the siege of 1016, in which an operation of circumvallation played a great part, as both the Chronicle and the Encomium (II, 7) emphasise. This is possible : Sigvatr was not present in England with Óláfr, and he may have worked on confused accounts of the operations at London, which were influenced by the events of 1016. We know that the siege of 1016 attracted much attention in Europe (see above, p. lx), and it would tend to obliterate or obscure popular memory of the nature of the earlier siege.

^{page 77 note 5} Ed. Thorpe, i. 162.

^{page 78 note 1} Although Óttarr's verse on Æthelred's return does not name Óláfr, there can be no doubt that it refers to him. If it were a genuine verse, but referred to some other person, no one would have thought that it referred to Óláfr, who was universally and truly believed to have fought on the Danish side in England (see below). On the other hand, it cannot be a forgery : no one would have dreamed of fabricating a verse depicting Óláfr as a friend of Æthelred. It must be an early verse, which was well known to refer to Óláfr, and hence caused saga-writers, who knew Óláfr was an enemy of the English, endless difficulty. The Hisioria Norvegiae (Storm, p. 124) says that Óláfr took four bishops with him to Norway when he left England. This, if true, would show that his last visit to England was friendly in the extreme, but the statement is probably no more than an unjustifiable inference from Adam of Bremen's account (ii. 55) of how English bishops worked for Óláfr.

^{page 78 note 2} v. 11–12. William makes it abundantly clear that it is Óláfr, the future king and martyr, to whom he refers, and alleges that his baptism took place in Normandy on this occasion. In view of the statement of the Oldest Saga, that Óláfr spent a winter by the Seine in the course of his European wanderings (see below, p. 81, note 9), and of Óttarr's statement that he brought Æthelred, whose place of exile was certainly Normandy, back to England, there is every reason to believe William's statement that Óláfr had been in Normandy. William says that this was at the time of Duke Richard's war with Odo of Chartres, and, although the evidence for the date of this war is not good, there is no objection to placing it 1013–14 (see F. Lot, Fidèles ou vassaux, p. 143), so that it would appear that William is also right as to the time of Óláfr's visit to Normandy. His account of the activities of Óláfr in Normandy is, however, strange. Óláfr and a totally unknown Lacman, called king of the Swedes, are invited by Richard to help him in his war. They hasten to his assistance (William clearly thinks of them as coming from their Northern realms), and, on their arrival in France, destroy Dol. The kings are nevertheless received with delight by Richard, who, however, does not require their active assistance, as the king of France intervenes and stops the war, when he hears of the barbarities indulged in by Richard's new allies at Dol. Now this is an absurd story, for the Bretons were fighting for Richard against Odo, yet, when Richard acquires new allies, they immediately destroy a Breton town, and are, nevertheless, joyfully received by Richard. Attempts to explain Óláfr's attack on Dol are made by Freeman, N.C., i. 460–1, De la Borderie, Histoire de Bretagne, iii. 3, Steenstrup, Normandiets Historie, pp. 163 ff. It would seem likely that the attack on Dol preceded the duke's invitation to the vikings to assist him, and that William's idea that it followed it arose from his belief that the Northern kings were at home in their own kingdoms when Richard's messengers approached them, and that they then set out for France. Actually they were doubtless ravaging up and down the coast: at least Sigvatr's poem shows that Óláfr was so engaged about this time, and evidence for viking activity on the French coast near Brittany shortly before Knútr became king of England is provided by the contemporary chronicler Ademar, M.G.H., SS., iv. 136 and 139–40, passages which are discussed by Steenstrup, loc. cit.

^{page 79 note 1} See below, p. 81.

^{page 79 note 2} Not apparently in England, where there do not seem to have been any traditional memories of Óláfr, for the reference to him in the so-called Loans of Eadweard the Confessor (Hoveden, Rolls Series, ii. 240) is obviously derived from William of Jumièges ; also not in the Celtic lands, where the memory of the warrior saint degenerated till it retained no trace of historical place or time (see Revue Celtique, xlii. 336 ff.).

^{page 79 note 3} v. 8. This incident is placed shortly before Óláfr goes to assist the Normans in the war discussed above.

^{page 79 note 4} ii. 49. Adam believed that Óláfr was a son of Óláfr Tryggvason.

^{page 79 note 5} Ed. Holder, pp. 343–4.

^{page 79 note 6} Gertz, i. 20. The passage is quoted, N.C., i. 704.

^{page 79 note 7} Storm, pp. 121 ff. This version is perhaps merely a literary combination of the versions of Adam and William of Jumièges. Historia Norvegiae also uses Óttarr's poem. Theodricus knew nothing of Óláfr's part in the fighting in England, except that he knew Óttarr's verse on the restoration of Æthelred. He was quite ignorant of the circumstances under which Æthelred went into exile, and has to fabricate an explanation, saying that Óláfr reconciliauit Adalŗedum fratribus suis et ut in regem sublimaretur obtinuit (Storm, p. 25). Ágrip is silent on the whole matter.

^{page 79 note 3} It is one of the most fixed elements in the northern chronology that Óláfr reigned fifteen years, but there is some doubt as to the point from which these were reckoned. If they are reckoned from his arrival in Norway, this must be placed in 1015, but if the first winter, before the defeat of Earl Sveinn, or the period after his flight to Russia in 1028, in which he was a king without power, be excluded from his reign, his arrival must be put in 1014. In the present work, I adopt the latter date, as slightly the more difficult for my argument. If the 1015 date be accepted, it becomes entirely out of the question that Óláfr helped Knútr in his invasion. The early career of Óláfr is carefully discussed in B. K. Brynildsen's Om tidsregningen i Olav den, helliges historie, and in O. A. Johnsen's Olav Haraldssons ungdom indtil slaget ved Nesjar (both Christiania, 1916). The former scholar accepts 1014 as the date of Óláfr's arrival in Normandy, the latter 1015, and I do not consider it possible to decide finally between these years. On the other hand, the fifteen years reign of Óláfr is a firm tradition, and is confirmed by a verse of Sigvatr (Skjaldedigtning, IA, p. 262 ; IB, p. 244), so it is manifestly impossible to make the period from the summer of 1030 back to the autumn of his arrival in Norway include less than fifteen winters; also, since he was present at Æthelred's restoration in 1014, it cannot include more than sixteen winters.

^{page 80 note 1} J. Schreiner : references to his works and criticism of them in Bjarni Aðalbjarnarson's Om de norske kongers sagaer, pp. 177 ff.

^{page 80 note 2} Óttarr's verse begins : Komt í land ok lendir ‥. Aðalráði; the Legendary Saga says that men say that Óláfr hafðe komet Aðalrað kononge aptr i land.

^{page 80 note 3} One of the only two references to Eiríkr in the Legendary Saga : see above, p. 71, note 1.

^{page 80 note 4} Nordal's work is referred to in the Preface.

^{page 81 note 1} The fragments of the Oldest Saga are edited by Storm, Otte Brudstykker af den ældste Saga om Olav den Hellige (Christiania, 1893). The Legendary Saga is edited by Keyser, and Unger, , Olafs saga hins helga (Christiania, 1849)Google Scholar, and by O. A. Johnsen (same title and place, 1922). The Oldest Saga is to be dated 1160–85. The Legendary Saga is the result of a curiously complicated evolution from the Oldest Saga (through M), concerning which Seip, D., Den legendariske Olavssaga og Fagrskinna (Oslo, 1929)Google Scholar, should be consulted, as well as Nordal. It survives in a thirteenth-century manuscript.

^{page 81 note 2} They are printed in Flateyjarbók, iii. 237–48.

^{page 81 note 3} Snorri's separate version is printed by Munch, and Unger, , Saga Olafs konungs ens helga (Christiania, 1853)Google Scholar ; a new edition (same title and place) by O. A. Johnsen and Jón Helgason has appeared (1941).

^{page 81 note 4} These compilations are analysed by Nordal. The most elaborate is printed in Flateyjarbók, ii, and a simpler one in Fornmanna Sógur, iv–v.

^{page 81 note 5} To whom it is to be credited is uncertain. Styrmir perhaps followed the Oldest Saga fairly closely on Óláfr in England, and Fagrskinna revised him without mercy. The general similarity of Snorri's story to that of Fagrskinna would then be due to the fact that Snorri, while basing his work on Styrmir, consulted Fagrskinnȧ. (Nordal leaves the question whether Snorri used Fagrskinna open, but Bjarni Aðalbjarnarson has proved that he did so, Om de norske kongers sagaer, pp. 173 ff.)

^{page 81 note 6} Thorkeil is not mentioned at all in Snorri's separate version, nor in Fagrskinna.

^{page 81 note 7} Note the removal of the meeting with Thorkell and the battle of Suðrvik from England to Denmark ; cf. above, p. 73.

^{page 81 note 8} From this point Fagrskinna is a bare summary, adding little to what the verses tell; I follow Heimskringla.

^{page 81 note 9} In the course of Óláfr's continental wanderings in the Oldest Saga, he spent a winter by the Seine, and the Legendary Saga and Fagrskinna repeat this. This is the starting-point of Snorri's story, but when he makes Óláfr be well received in Normandy, and meet the sons of Æthelred there, he is neatly combining two facts recorded by William of Jumièges, firstly that Óláfr was at one time an ally of Duke Richard and secondly that the sons of Æthelred fled to Normandy during the Danish invasions (vi. 10). William's history was known in the North, and there is not the least difficulty in assuming that Snorri knew it directly or indirectly.

^{page 82 note 1} See below, p. 91.

^{page 82 note 2} On the various forms of the name, see Napier and Stevenson, Crawford Collection, pp. 139 and 142.

^{page 82 note 3} For example, E, F, and Henry of Huntingdon place Knútr's death in 1036 (C, D, 1035).

^{page 82 note 4} Stenton, p. 397, places the battle in 1026, going against both the English and the Scandinavian evidence without giving his reasons. It is well known that, when he was at Rome early in 1027, Knútr addressed a letter to his people, in which he says that he is about to go to Scandinavia to deal with a movement of certain peoples against him (Florence of Worcester, ed. Thorpe, i. 188 ; cf. above, p. lxii). This almost certainly refers to the Swedish-Norwegian alliance, which he was facing in the campaign in which the battle of Helge-ȧ occurred.

^{page 82 note 5} Ed. Holder, p. 348.

^{page 83 note 1} Individual sources make small independent additions : Fagrskinna and Heimskringla revising the early Saga of Magnús Góthi, which we know from Morkinskinna and Flateyjarbók, axe able to add his place of burial; Heimskringla also gives the length of his reign in England correctly. Various scraps of information are also found in Knytlinga Saga.

^{page 83 note 2} It occurs in Fagrskinna, pp. 183–91. One passage of it is already present in the Legendary Saga, chap. 100, but it was not in the Oldest Saga, as the Fragments show. Nordal (op. cit., pp. 162–3) is not convinced that a separate Saga of Knútr is the source of this material, but cf. Indrebø, G., Fagrskinna (Christiania, 1917), pp. 101–3Google Scholar, where the best discussion of the question will be found. The only point which concerns the present enquiry is that we have here an addition to the Saga of Ólá;fr upon the affairs of the Danish royal house.

^{page 83 note 3} P. 161.

^{page 83 note 4} P. 183.

^{page 84 note 1} Rather similarly Saxo places the battle of Helge-å after Óláfr's flight to, and return from, Russia.

^{page 84 note 2} Óláfs Saga Helga, chaps. 148–53.

^{page 84 note 3} Snorri quotes a good deal of verse about the campaign, especially from two poems (both in praise of Knútr, and both called Knúitsdrápa) by Sigvatr and Óttarr. He would reasonably think that, if Knútr had had any important adversary besides the kings of Norway and Sweden, one of the poets would have mentioned it. Modification of a narrative in conformity with the silence of skaldic verse is not an unknown process in the development of the Sagas: for an interesting instance, see Storm, Snorre Sturlassöns historieskrivning, pp. 143–4.

^{page 84 note 4} The verse is the first of the Glælognskviða of Thórarinn Loftunga (Skjaldedigtning, IA, p. 324; IB, p. 300). It is quoted by Fagrskinna (p. 183) to illustrate the statement in the insertion from Knúts Saga that Úlfr accompanied Sveinn, and in Heimshringla to support the statement that Haraldr did so (see below).

^{page 84 note 5} Óláfs Saga Helga, chaps. 183 and 239.

^{page 84 note 6} I have drawn attention to examples of this in the Viking Society's Saga Book, xii. 232–7.

^{page 84 note 7} See above, p. 75.

^{page 84 note 8} The Encomiast's desire to depict Thorkell as loyal to the Danish interest at all times may well be due to the fact that at the time he was writing Thorkell's son was alive and married to a near relative of the king (cf. above, pp. 74–5). It is interesting that Snorri knew that Thorkell had a son called Haraldr, although he played no noteworthy part in history. The survival of scraps of genealogical information in the North is often surprising: an example is Snorri's knowledge of the existence of Æthelred's two obscure sons Eadwig and Eadgar (Heimskringla, Óláfs Saga Helga, chap. 20).

^{page 85 note 1} Cf. N.C., ii. 65, n. 3, where it is suggested that these children of Gunnhildr might be children of her first husband: but her first marriage is doubtful (see below), and the names of her children make it certain that they belonged to Thorkell's family.

^{page 85 note 2} It may be noted that the passage in Knytlinga Saga, chap. 75, where Hakon Eirlksson is alleged to have had a daughter is historically worthless. The Worcester Cartulary agrees with Florence that his wife was called Gunnhildr (see N.C., ii. 579–80).

^{page 85 note 3} This Haraldr's signature also appears in the forgery K. 1327 (= R. 85). K. 749 is signed by a Haraldr, but he is not described as dux, and many unknown persons with Scandinavian names sign this charter.

^{page 85 note 4} See Adam of Bremen, ii. 75, and Stenton, pp. 417–18, for the circumstances. The year, however, was 1042, not 1043 (see Steindorff's Jahrbücher des deutschen Seiches unter Heinrich III, i. 275, footnote 1, and further literature there quoted). The day is known from the Necrology of St. Michael's, Lüneburg, which enters under 13 Nov., obiit Haraldus dux et occisus. This is a good authority, for this Necrology incorporates early material, and St. Michael's had close connections with the Danish royal house in the eleventh century (see Aarbøger for nordisk oldkyndighed og historie, 1927, p. 31, footnote).

^{page 85 note 5} See Knytlinga Saga, chap. 11.

^{page 85 note 6} Ed. Thorpe, i. 201–2 (clear though by implication); ibid., p. 275, and ii. 2, calls her a sister of Sveinn Úlfsson of Denmark: she was, of course, his aunt.

^{page 85 note 7} Fagrskinna, p. 279; Heimskringla, Óláfs Saga Helga, chap. 152; Knytlinga Saga, chap. 11.

^{page 85 note 8} Ed. Holder, p. 350; but cf. N.C., i. 744.

^{page 85 note 9} Chap. 11.

^{page 86 note 1} A rather fuller discusssion of Ástríthr's marriages than that of Freeman will be found in K. Maurer, Die Bekehrung des norwegischen Stammes, i. 472–3, n. 24; Steenstrup, Normandiets Historie, pp. 226–7, may also be consulted, though his attempt to place Ástríthr's Norman marriage in the time of Sveinn's negotiations with Normandy (see above, p. xlii, note 4) is not to be supported.

^{page 86 note 2} Reference as above, p. 70, note 6.

^{page 86 note 3} See N.C., i. 791–2; and, for an attempt seriously to connect Siward with Úlfr's family, Steenstrup, Normannerne, iii. 437 ff.

^{page 86 note 4} Anglo-Saxon Bishops, Kings and Nobles, p. 355.

^{page 86 note 5} Scriptores, iii. 281–2.

^{page 86 note 6} Saxo does not make it clear whether Óláfr of Norway takes part in the battle in his version of the story. H. Koht., Inhogg og Utsyn, pp. 136 ff., attempts to prove on insufficient grounds that he did not; cf. Schreiner, J., (Norsk) Historisk Tidsskrift, xxvii (1927), pp. 311 ffGoogle Scholar.

^{page 86 note 7} N.C., i. 765.

^{page 86 note 8} See Napier and Stevenson, Crawford Charters, p. 142, for references.

^{page 86 note 9} Also the notorious forgery K. 1327 (= R. 85).

^{page 86 note 10} Reference as above, p. 70, note 6.

^{page 86 note 11} This event is noticed by the only two manuscripts of the Welsh Latin annals which have entries in the period, and by the Welsh vernacular chronicles.

^{page 87 note 1} W. J. Rees, Lives of the Cambro British Saints, p. 77.

^{page 87 note 2} The notice does not appear in the Welsh Latin annals, but there is every reason to regard it as an early one for it appears both in Welsh chronicles of the Red Book type, and in MS. Peniarth 20 in identical words: the death of Knútr is recorded, and it is then said that gwedy y varw efy foes Eilaf hyt yn Germania. In the early Welsh annals Germania often means Norway : Knútr is described as king of Germany, Denmark and England, and Harðráði and Magnús Berfoetr are both called kings of Germany. On the other hand, the word can have its usual meaning also.

^{page 87 note 3} See above, p. 73.

^{page 87 note 4} Suhm, Historie af Danmark, iii. 502 (followed by various authorities), sees in Wolf, the mythical ancestor of Wigbert of Thuringia, confused memories of Earl Úlfr (see M.G.H., SS., xvi. 234 ff.). This is mere nonsense, not worth discussion.

^{page 87 note 5} See Finnur Jónsson, Den oldnorske og oldislandske litteraturs historie, ii (2nd ed.), p. 778.

^{page 87 note 6} See Crawford Collection, p. 140, note 2.

^{page 88 note 1} On the relationships of these Sagas, see above, pp. 80–1.

^{page 88 note 2} Cf. above, pp. 70-r and 80.

^{page 88 note 3} It is particularly noteworthy that the Supplement gives Eiríkr as an alternative name for Eadric Streona, while preferring to use the form Alrekr : the use of the name Eiríkr for this man is peculiar to the Legendary Saga, where it, no doubt, comes from the Oldest Saga.

^{page 89 note 1} See above, p. 73.

^{page 89 note 2} N.C., i. 670.

^{page 89 note 3} See above, p. 75.

^{page 90 note 1} See N.C., i, 648 ff. If the massacre described in the Supplement is, in fact, founded, upon no more than vague memories of some of the incidents of 1002, the representation of Knútr's invasion as a mission of vengeance may arise from the fact that the invasion of Sveinn, which followed upon the massacre of 1002, was in some quarters believed to have had personal vengeance as its object. (See Stenton, p. 375, where William of Malmesbury's confused statement, Gesta Regum, ii. 177, which literally means that Sveinn's mission of vengeance was the 1013 expedition, is no doubt correctly applied to that of 1003. William of Jumièges, v. 6, also attributes Sveinn's expedition of 1003 to a desire to avenge the massacre of 1002, but Adam of Bremen, ii. 49, makes him wish to avenge a brother, presumably the obscure Hiring, on whom see my Battle of Brunanburh, pp. 71–2. William and Adam both telescope Sveinn's expeditions of 1003 and 1013).

^{page 90 note 2} Langebek, Scriptores, ii. 459, sees the chronological impossibility of Thorkell's brother, who was killed before the death of Sveinn, being in charge of the thingmen after that event, and he solves the problem by bringing Thorkell's other brother Sigvaldi to England. Sigvaldi is killed and avenged before Sveinn's death, Hemingr afterwards. This piece of perverted ingenuity is reproduced by various respected authorities on the history of the period. There is, of course, no shadow of evidence that Sigvaldi was ever in England.

^{page 90 note 3} Similar stories will be found referred to in the controversy about the story of Sigurthr Slefa between Boer, R. C. and Jónsson, Jón in Arkiv för nordisk filologi, xviii. 97Google Scholar; xxvi. 202 and 346; xxvii. 192.

^{page 90 note 4} See Nordal, Oṁ Olaf den helliges saga, p. 118, on the purely literary reasons which led to the fabrication of this character (probably by Styrmir).

^{page 91 note 1} Those who wish can study this production in Miss Ashdown's English and Norse Documents, pp. 176 ff.