The present translation of the shorter version of Astute readers will note that I use og instead of ok in the title, preferring to standardize to modern Icelandic rather than use an Old Norse orthography. My reasons for doing so are explained briefly in the note preceding the translation. For consistency’s sake, I always use modern Icelandic orthography (rather than switching between two standards), even when discussing characters that many will consider to be medieval. This is because the discussion here rests principally on texts produced in the sixteenth and seventeenth centuries, during which period Icelandic scribes themselves often wrote names in a non-medieval form. In practice, this means that as well as using og instead of ok in titles, I also use an epenthetic u in the endings of names such as Konráður and Hrólfur. (hereafter simply ) with accompanying Old Norse text and introduction is one more contribution to a series of similar English translations that have appeared in recent years (O’Connor and O’Connor; Hall et al. 2010; Hall, Richardson, and Þorgeirsson 2013; Lavender 2015; McDonald Werronen 2016, 235–48 [rev. from McDonald 2009]).It is worth noting that these more recent translations have been inspired by earlier work connected with Leeds University, such as Keren H. Wickʼs work on 1996, Alenka Divjakʼs work on 2009, and Andrew Wawnʼs work on 2010. The general aim of the other contributions, to make accessible and encourage study of native Icelandic romance through detailed case-studies of individual sagas Hall et al. 2013, 81; see also Hall 2014, remains one of the principal motivations for the work that has been undertaken on . In the introduction that follows, as well as relevant information on manuscripts and dating, particular discussion will be provided on the subject of the different versions of the saga, the principal motifs, the question of homosocial relations, and the reception of the saga in the post-medieval period.

In their Marianne Kalinke and P. M. Mitchell list (under the title of ) 71 manuscripts containing some form of spread over public and private collections in Denmark, Iceland, Norway, Sweden, and the US.On the different titles see below, footnote 21. Four of these 71 witnesses (AM 576 b 4to, Lbs 3128 4to, NKS 1144 fol., and NKS 1685 b I 4to) contain a resumé or summary of the narrative, rather than a full text. A further three (Lbs 360 4to, Lbs 2233 8vo, and Rask 77) are known to be defective or fragments. One of the manuscripts, Säfstaholms​samlingen I Papp. 11 at Riksarkivet in Stockholm, also contains a translation of the saga into Swedish. Another of the manuscripts listed by Kalinke and Mitchell, ÍB 151 4to, does not actually contain the saga: it appears to have been included on their list in error. Since that time six additional manuscripts containing the saga have been catalogued, five of which now appear on , giving, once we exclude an erroneous entry from Kalinke and Mitchell’s list, a total of 76.Those on handrit.is are: Lbs 5157 4to (1899–1903); Lbs 5186 4to (1899–1903); Lbs 4729 8vo (1903); SÁM 6 (1800–1900); and SÁM 74 (1888–1889). The manuscript not yet on handrit.is is ISDA JB6 1 (1825–1851) at the Elizabeth Dafoe Library in Winnipeg. I have not included AM 591 e 4to among the 77 manuscripts, although the lower half of f. 19v in that manuscript apparently contained the start of , before it was scribbled out. See Loth 1960, 134–35. Even during the preparation of this article new witnesses have come to light, so it is almost certain that once Katelin Parson’s findings and catalogue based on her research in Canada are made public the list of texts of will have grown even longer.

The oldest extant manuscript containing the saga is AM 556 b 4to, which is also the principal text of the longer (or older) version, listed as being from between 1475 and 1500 in the database of the (ONP).Kålund (1889–1894, I:707) listed the manuscript as being simply fifteenth-century. Handrit.is agrees with the ONPʼs dating. ONP datings are used for all other manuscripts here and are generally in agreement with those appearing on handrit.is. The oldest manuscript of the shorter (or younger) version, AM 529 4to, has been dated between 1500 and 1550. One more sixteenth-century manuscript of the longer version, AM 510 4to from around 1550, exists as well as between four and five from the seventeenth century.A manuscript in Bergen, UBB MS 1491, is dated to c. 1700, thus on the cusp. Árni Magnússon’s summary in AM 576 b 4to is dated, similarly, to between 1690 and 1710. That means that only seven or eight (approximately 10%) of all the extant manuscripts are from before the year 1700. The vast majority are eighteenth- or nineteenth-century copies.

Based purely on the extant manuscripts, we could say that the longer version of the saga came into being in the second half of the fifteenth century and the shorter one, an offshoot of its more prolix companion, did so shortly afterwards, probably in the first half of the sixteenth century. Such a conclusion, while reasonable, is, however, by no means inexorable. The saga could have its roots further back than its earliest extant witnesses, and it may be mere coincidence that the shorter version happened to be preserved for posterity in a younger manuscript than the longer version. Nevertheless, a comparison of similarities (motifs and plot elements) with other texts suggests that until further proof comes to light we would do well not to diverge too far from the most straightforward (albeit provisional) hypothesis.

Kalinke 1990, 169; 2017, 113, adopting Gustaf Cederschiöld’s reasoning in 1884, cliii, sees as a response to .See also Bibire (68) and the section on homosociality for further discussion of the relationship between these two sagas. It is interesting to note that if was a response to , it may be considered to have outshined its predecessor, which exists in only 48 manuscripts , xvii. The reference to Konráður at the end of the shorter version of could be seen as a way of making explicit the connections with his story already implicit in the longer version, and since the earliest extant manuscript of is Perg. 4to nr 7 (from the Kungliga Biblioteket in Stockholm), dated 1300–1325, it is natural to assume that is the more senior partner.It is of course possible that an earlier form of was not intended to invoke associations with Konráður’s narrative, these being worked in after the latter’s saga became popular. Since this would have taken place for the most part prior to the writing of any of the extant manuscripts, however, it must remain pure speculation. The presence of a prologue stating that the story was found written on a stone wall (the graffiti claim; see Barnes 271) in the longer version of has also been adduced as proof of the influence of , since the latter is also said to have been found written on a wall. If we accept this vector of transmission, then it is even more probable that the longer version (the only one to contain the graffiti motif) must be the older, and that this section came to be omitted subsequently in the shorter version. In a comparison of motifs and textual features shared across a number of texts, Hall et al. corroborate the placing of all these jigsaw pieces with their insertion of after (dated late thirteenth–early fourteenth century) and before and (both dated fourteenth–mid-fifteenth century) 2010, 67–70. Other similarities with , such as the twelve men on horseback with one in the lead larger than all the others or the heathen threat of taking a princess either as a wife or a concubine, have been noted by Astrid Van Nahl (33–35) and Henric Bagerius (256) and thus strengthen the grounds for assuming a chain of influence.Interestingly, Astrid van Nahl implies that the influence is moving in the opposite direction, when she states that a scene in deutlich an die Vilhjálms saga sjóðs errinert 34. The trail of the graffiti saga, however, leads us to believe that the inverse is more likely. It should be mentioned that the reminiscences of Tristan in the form of the proxy wooer (arousing jealousy, justly or not) and the disguise that Jarlmann adopts to approach Ríkilát, while patent, are no help in dating: was one of the earliest romance translations to be carried out in Scandinavia (if not the earliest) and its influence is pervasive Schach, especially 91.

As for the place of composition, Stefán Einarsson (272) claimed that the saga came from Oddi in southern Iceland, but provided no explanation as to why he believed this to be the case, nor is there any obvious local flavour in the saga that would connect it to a specific part of Iceland. None of the scribes or commissioners of the early manuscripts are known.

There are two reworkings of into the long poetic narrative form known as rímur, both of which are later than the earliest extant manuscripts and so add nothing new with regard to refining the dating. The first, , is by an unknown author and exists in full in two manuscripts, AM 610 c 4to and AM Acc. 22, and partially, up until midway through the eleventh ríma (or fitt), in AM 604 f 4to. Finnur Sigmundsson dates these rímur no more specifically than pre-1600, but Haukur Þorgeirsson (256) has recently assigned them on metrical and linguistic grounds to the period between 1500 and 1550. They have never been edited. The second reworking is called , composed by Guðmundur Bergþórsson (1657–1705) and found in ten manuscripts. These too have never been published in full, but extracts from the second, eleventh, and fifteenth rímur (or fitts) appear in Sir William Craigie’s anthology of rímur verse (II:108–27).

In AM 576 b 4to on f. 12r a number of notes appear in Árni Magnússon’s hand under the heading Ur Sỏgu Jarlmanns oc Hermanns from the Saga of Jarlmann and Hermann. After being told that prologus quibusdam deest the prologue is absent in some (versions/manuscripts), Árni explains how the saga starts: Mestare Virgilius hefe samsett marga fræði til skiemtanar mönnum i bok þeirri er Saxa fræði heitte enn þessa sỏgu fann hann skrifada a steinvegginum borgar þeirrar er lisibon heitte i frans. Vilhialmur enn aurfi (quibusdam dictus) kongur i fracklandi (alii addunt Gardariki og þiskalandi) atti dottr kongsins af lumbardi þeirra bỏrn voru Hermann er riki tok eftter fỏdur sinn og Herborg. Jarlmann het fostbroder Hermans son Rodgeirs Jarls af Granuckatis i Frack landi (alii Saxlandi). Master Virgil is the author of a great deal of learned information that entertains men in the book named “Lore of the Saxons,” and he found this story written on a stone wall in the town named Lisbon in France. Vilhjálmur the Generous [as said in some texts], king of France [others add Russia and Germany] was married to the daughter of the king of Lombardy. Their children were Hermann, who inherited the kingdom after his father, and Herborg. Jarlmann was the sworn brother of Hermann, son of Earl Roðgeir of Granuckatis in France (others say Saxony).This translation, as all others unless otherwise stated, is my own. The notes continue in a similar manner, but from the extract shown above and its parenthetical comments it is already clear that in the seventeenth century Árni was comparing different versions of the text. What is also clear is that the major differences with which Árni concerned himself (beyond the presence or absence of the prologue) were the different personal names or locations mentioned.

Three centuries later Björn K. Þórólfsson, while looking at , also made observations about the two versions of : Sagan, sem Jarlmannsrímur eru kveðnar eftir, Jarlmanns saga og Hermanns, er til í tveim gerður, og er önnur gerðin eldri og betri en hin töluvert breytt. Hugo Rydberg gaf söguna út 1917 eftir handritum af yngra flokknum, skinnbókinni AM 529, 4to og pappírshandritinu AM 167, fol. En hin eldri gerð sögunnar er í skinnbókunum AM 510, 4to og AM 556 b, 4to, og eftir þeirri gerð eru rímurnar kveðnar. Þær segja, eins og eldri gerðin, að Ríkilát keisaradóttir hefði krossin, sem hún var fædd með og læknisgáfan fylgdi, á hægra hendi, og er það auðvitað upphaflegra en hit, sem segir í yngri gerðinni, að krossinn væri á vinstri hendi Ríkilátar. Af fylgd Ríkilátar, er hún fór til kirkju, segir miklu meira í eldri gerðinni og rímunum en yngri gerðin hermir. Það er auðsjeð, að yngri gerðin er stytt úr hinni eldri, og má því ætla, að upphafleg sje sú frásögn, sem eldri gerðin og rímurnar hefjast á, en ekki finst í yngri gerðinni, að Virgilíus fyndi Jarlmannssögu á steinnveggjum (marmara í rímunum) í Lissabón. 452–53 The saga on which are based, , exists in two versions, one of which is older and better and the other of which is significantly altered. Hugo Rydberg published an edition of the saga in 1917 based on manuscripts of the younger version, namely the parchment AM 529 4to and the paper manuscript AM 167 fol. The older version of the saga is in the parchments AM 510 4to and AM 556 b 4to, and the rímur are based on this version. They, like the older version, recount how Ríkilát, daughter of the emperor, had a cross, which she had been born with and which had healing powers, on her right hand, and that is clearly more original than what is recounted in the younger version, that the cross was on Ríkilát’s left hand. On the subject of Ríkilát’s retinue, when she goes to the church, much more is said in the older version and the rímur than the younger version includes. It is clear that the younger version is an abbreviation of the older one, and it may be assumed that the story with which the older version and the rímur begin, that Virgilíus found on a stone wall (marble in the rímur) in Lisbon, is original. Björn’s statements cover new ground in recognizing that the changes in the different versions of the saga go beyond mere naming conventions, but there is room for debate as regards some of them. Firstly, the question of whether the older (or longer) version is better could perhaps be replaced with the statement that the two versions function in somewhat different ways and may be considered better or worse in differing contexts, i.e. when specific audiences apply varying criteria to them. As an example, the translation of the shorter version provided here was motivated in part by the fact that Agnete Loth, with the help of Gillian Fellows Jensen, had already provided a paraphrase of the longer one, but also by the desire to share the slightly more fast-paced and thus potentially more engrossing shorter version with a wider audience of both general and specialist readers.

Secondly, the comments on age and the trajectory of textual development between the two versions can be challenged. While not clearly wrong, they are at the same time by no means definitive proof of the two versionsʼ relationship, and, accordingly, Marianne Kalinke has recently stated that she does not believe that the older redaction represents the author’s composition, but rather that both redactions derive from a text no longer extant 2017, 120. Taking the second of Björn K Þórólfsson’s points first, the more detailed description of Ríkilátʼs retinue could surely be an expanded form of an earlier and shorter description, as in the shorter version: prolixity is no unambiguous sign of greater age. The first point Björn K. Þórólfsson makes, namely that Ríkilát’s cross is on her right hand in the older version (and thus that version is clearly more original), but on her left hand in the younger version, is also ambiguous. Birthmarks or scars in the shape of a cross are a typical sign of royalty and/or sanctity in medieval literature. Inger M. Boberg’s motif index refers, however, to only two examples of something similar under H71.5 Marks of Royalty, those being the case under discussion and that of Þorleifur breiðskeggur in 175. Þorleifur, a pretender to the throne following the death of King Magnús Erlingsson in 1184, is said to have been son Eysteins konungs Haraldssonar, ok þat til jartegna at á meðal herða honom var ørr gróit í kross the son of King Eysteinn Haraldsson and as a sign thereof bore a scar between his shoulders that had healed in the shape of a cross. Haki Antonsson has commented that the stories of Þorleifur and his scar are one example of attempts by followers of ill-fated challengers to royal power to cultivate the saintly reputation of their leaders 176. He also points out, referring to Marc Bloch’s study , that similar signs (cross-shaped birthmarks), if not found between the shoulders, were usually found on the right shoulder of the person in question 178. There is no great overlap between Ríkilát’s and Þorleifur’s markings and social situation (birthmark/scar, rightful princess/failed pretender), and these two examples alone are not sufficient to claim that their appearance was a familiar motif in Iceland and followed a recognized set of rules. If we accept that the right-side orientation is a more orthodox positioning of such signs, then the shorter version’s shift to the left could be read as an unintentional alteration made by a scribe unaware of the valency of the right-side orientation. The alternative, that this was a purposeful subversion of the motif, should also be considered, but there is no obvious reason why a scribe should wish to subtly call Ríkilátʼs religious status into doubt. Whatever the case, this small difference should not be read as proof that one version is definitively anterior to the other. Both sagas contain idiosyncratic elements that may be read as more or less traditional, and more examples than the descriptions of the retinue and the cross-birthmark alone would need to be adduced to prove that one or the other version was more original.

Thirdly, but also connected to the previous point, we must deal with the idea of abbreviation. The shorter version is—self-evidently—shorter, but not all the differences consist of omissions or abbreviations. There are also details not present in the longer version, such as when Jarlmann draws a picture on the wall for Ríkilát so that she may know what Hermann looks like. As Marianne Kalinke 2017, 120–21 has argued, this episode, in her view derived from and drawing on scenes where an individual falls in love by seeing an image of a person (as in Tristan stories), could well be a more original feature of the saga that has been lost in the longer version. It may even, before its hypothetical loss, have encouraged the inclusion of the prologue in the longer version: if Hermann can be inscribed upon a wall visually, why not his story textually? The two versions also show rearrangements of material, such as in the run-up to the conflict in Miklagarður (Constantinople). In the shorter version Rómanus’ men confront the King, after which Jarlmann arrives and offers his help. In the longer version Jarlmann goes to offer his help before the bullying envoys arrive to make explicit their demands for Ríkilát. Quite why this is the case remains to be determined, but it provides yet more proof that the difference between the versions goes beyond mere abbreviation.

Ultimately, it seems safest to refrain from rash proclamations about the relationship of the two versions until further work has been carried out in comparing the major manuscripts and producing a stemma. Some provisional work, which it is hoped will be published in the near future, has been carried out involving the comparison of small sections of text from the various witnesses of the saga. In the meantime, however, an appendix to this introduction provides a table which allows for easy comparison of the shorter version of the saga with the longer version (with chapter numbers) and highlights all the significant differences in italics.

The names of the characters in our saga provide a starting point for tracing its interconnectedness with other works. Many of them are exotic inventions (Ríkilát) or typical placeholders (Rúdent) for foreign personages.Compare for example Ríkiza in (Loth 1962–1965, IV:185) and Rúdent from Serkland in (Loth 1962–1965, V:177). Ermanus and Rómanus, for example, are almost interchangeable (as shown by the fact that they are swapped around in the two versions). Hermann and Jarlmann seem to function almost allegorically. Her means army, with a hermaður being an individual within an army or a warrior, and the accusative form (hermann) contributing to Old Norse compounds, such as hermannligr (warlike or gallant). Jarl means earl, and jarlmaður (accusative jarlmann, literally earl-man) is more or less a synonym. Thus we have the bellicose ruler and his obedient subordinate,Hermann is not an extremely common name in saga literature, but a character named thus also appears in and (see Lavender 2016). who are ostentatiously of different social ranks (hence one is subjected to the whims of the other). Nevertheless, their roles are complementary: their adopted names of Austvestan (East-west) and Norðsunnan (North-south) imply this. Þorbjörg, or its variant Þornbjörg, is also not a particularly common name in romance sagas. Only one Þor(n)björg appears elsewhere in fornaldarsögur, as the outstanding meykongur (maiden-king) who later becomes the hero’s wife (and saviour) in . It must be remembered, however, that, as Kalinke 1990, 26, 65 points out, this latter saga is paradigmatic for and provides a survey of the possibilities of bridal-quest narrative in Iceland. Among the riddarasögur, another maiden-king bears the name Þornbjörg in V:134. It was clearly a name fitting for prodigious women. The Kuflungar (or Cowled Ones, hereafter Kuflungs), the nickname of the evil characters who build a hall for Hermann, is a designation shared by a rebellious faction from late twelfth-century Norway mentioned in 155–56. As we have already seen in the case of the mark of the cross, displays certain elements that may have been taken over into . Yet in this case, as in that of the auspicious birthmark, the context and exploitation of an image or a word is vastly different between these two sagas. Cowls, hoods, and cloaks, moreover, are frequently a sign of subterfuge, danger, and unexpected plot twists across the spectrum of saga genres, including riddarasögur.One of Óðinn’s alternative names, for example, is said in to be Síðhöttr long-hood, presumably due to his tendency to conceal his face and thus identity before causing problems for someone Snorri Sturluson 22. Örvar-Oddur hides under a feldr cloak to avoid an unwished for prophecy by a prophetess , II:206–207. Many more examples could be given. Thus on the basis of the names alone fits squarely within the broad range of romantic and legendary sagas.

In her whirlwind tour of Icelandic romance literature, Margaret Schlauch mentions in connection with a number of the motifs she discusses: the image of the fair unknown arousing feelings of love 65; the abduction of a bride and subsequent rescue by her future bridegroom 66, 106; an irreproachable proxy suitor 152; a princess with the gift of healing 153; and an excellent king being reminded that the one thing he lacks is a fitting wife 164.Astrid van Nahl also provides a list of the motifs present in along with their appearances in other originale Riddarasögur 216–18. Perhaps the most telling motif she lists is that of a wooer disguising himself in humble or beggarʼs clothes (as Jarlmann does in order to get near to Ríkilát), which only appears elsewhere in and . As Schlauch points out, some of these motifs are staples of romance literature, some too with their roots in Tristan traditions, and with such a wide dissemination that substantive conclusions on textual relations cannot be drawn from their mere appearance. There are several moments in , however, where a certain originality shines through, and it is worth considering these in greater detail. The two points that will be looked at here are the master-builder motif and that of the troll party.

shows, if anything, a little less interest in outlandish architecture and human interactions with it than many of its generic companions. In we encounter chambers with floating objects due to magnets embedded in walls, and in a hollowed-out mountain that can be filled and emptied using a system of hydraulics plays an important role.See also , which tells, among other things, of a floating house and the construction of lighthouses, the latter description said by Wick 226 to be another manifestation of ʼs interest in buildings and other works of construction. Nevertheless, a feel for the mystery and challenge of buildings remains. In chapter 9 the detail of the giant chain used to close off the sea-passage of the Golden Horn is casually alluded to, and in chapter 23 Hermann/Norðsunnan must escape Rúdent’s castle by jumping down from a towering height apparently off the top of the city wall. In this world suffused by elements of construction it becomes perhaps more believable that Hermann would be impressed and let his guard down when a group of skilled architect-builders arrive at his court, even if their attire should sound an alarm.Earl Svívari in is also fooled by some builders in disguise, who kidnap his charge, Princess Dorma Wick 131–36, 251–52. The short version of refers, moreover, to the shutting off of the Golden Horn, but in this case with an iron gate rather than a giant chain Wick 77, 80. It may thus be that this saga is inspired by .

Marianne Kalinke has argued that, by contrast with in particular, aligns itself more with Iceland’s indigenous folklore of giants, trolls, ogres, and elves 2017, 119. She has also pointed out that the arrival and subsequent actions of the Kuflungs may be considered a variation of the master-builder tale that has been transmitted to us in Snorri Sturlusson’s as well as in and 1990, 176, thus situating the episode in a distinctively Icelandic literary tradition. It is true that the coinciding features of the construction of a building within a specific time and the loss (potential or realized) of a woman are present in all these cases, but the treatment is varied. Unlike the giant who wishes to receive Freyja in payment, but is prevented at the last minute by the wily gods, the Kuflungs never reveal their intention to leave with Ríkilát and thus no defence can be mounted. This is perhaps the very point: had Hermann not sent Jarlmann away, then the latter would have perceived the danger and taken steps to neutralize the threat, just as Loki does. As it happens, on the completion of the plan not only has Ríkilát disappeared, but the hall proves structurally unsound, negating even this benefit. A detail found in both versions of the saga states that after the Kuflungs left hallargólfið [var] í sundur og kominn upp kolblár sjór the hall floor had cracked, and up through it had come coal-black seawater. This odd detail is reminiscent of various occasions in sagas where ships become damaged or sabotaged and thus spring a leak. In , for example, Án skaut einum fork, er járni var búinn, undir rangbyrðit á skútunni, ok gerðist þat skipti á, at sjór kolblár kom í staðinn Án launched a forked iron-plated spear at a point below the gunwale on the skiff, and it made a rent so that the coal-black sea poured in there , II:397. In chapter 11 of there is a similar example of a ship being purposely damaged to prevent pursuit , 35. In light of this, the fact that the hall springs a leak is fitting, since the Kuflungs also manage to evade all attempts at tracking them (just as if they had literally undermined their enemies’ means of conveyance), yet still the transposition of the seeping kolblár sjór from a ship to a hall is somewhat surreal and gives an uncanny edge to the whole episode.

Fittingly, in order to effect his own escape from Serkland, Hermann arranges for his enemies’ ships to be secretly sabotaged, so that when Rúdent’s men take chase, water pours in, and many of them drown. It seems he has learnt from his foes’ own ploys. It is also noteworthy that it is Ríkilát who ingeniously seals the fate of the master-builders. In a scene reminiscent of Ingigerður’s demands to Eiríkur at the start of , III:171–72, Ríkilát agrees to be pliable with her abductor as long as he fulfills certain conditions for her. One of these is to have the Kuflungs summarily executed. This may seem on the surface to be an act of bloodthirsty revenge, but can also be interpreted as a shrewd act of self-preservation (an interpretation that is espoused by one of Rúdentʼs gossiping knights). Ríkilát, aware that convention demands a rescue attempt, eliminates the only individuals at court who are familiar with Hermann and his circle of followers and so might expose her potential saviour when he arrives as Norðsunnan. In this sense, she surpasses Freyja in Snorri’s version and shows how women can also play a role in their own self-preservation.In this sense, Ríkilát can be compared to Nítíða in her eponymous saga, also a paragon of self-preservation McDonald Werronen 128, 169.

Another example of a formidable woman who is loath to relinquish agency is Þorbjörg the Stout, Rúdent’s oversized aunt. One of the most memorable scenes in the saga must surely be the party (or perhaps athletics meet) over which she presides at Mount Baldak/Kaldbak. In it we see, through the eyes of our human characters, what happens when otherworldly beings, including mikill fjöldi trölla a great throng of trolls as well as álfar og allt smáfólk elves and all little people get together and enjoy themselves. This scene allows us to get a peek into the types of events hinted at in other sagas. In , for example, the hero comes across a trollwoman one night, who tells him the following: Ek skal til tröllaþings. Þar kemr Skelkingr norðan ór Dumbshafi, konungr trölla, ok Ófóti ór Ófótansfirði ok Þorgerðr Hörgatröll ok aðrar stórvættir norðan ór landi I must go to the troll assembly. Skelkingur, king of the trolls, will attend, coming from Dumbshaf in the north, so too Ófóti from Ófóti’s fjord and Þorgerður Hörgatröll and other important beings from the north of the land II:173.Stephen Mitchell 87–88 has listed this example as one possible Icelandic portrayal of the growing interest in witchesʼ sabbaths in Scandinavia in the late medieval period. It could thus be the case that has also been influenced by these traditions. Of particular interest for this is the mention of the mountain (called both Baldak and Kaldbak), whose name is vaguely reminiscent of other infernal meeting places such as Blåkulla and Blaakolden Mitchell 83. We do not follow the trollwoman to the actual event in , but perhaps it would be something similar to the gathering witnessed in where þeir bræðr, Gautr ok Hildir, láta þing stefna. Kom þar til fólk um alla Jötunheima. Þar var ok Skröggr, því at hann var lögmaðr tröllana those brothers, Gautur and Hildir, summon people to the assembly. People gathered there from all over giantland. Skröggur was also there, because he was the lawman of the trolls , III:356. In this description the þing assembly is taken very literally as a place where the supernatural inhabitants of the world meet and make decisions, according to a set of rules. In other sagas we see trolls having unruly feasts in caves, but is the only example where so much singing, dancing, and sport is involved. The details of some of the entertainments are so convincing that Jón Samsonarson has stated that að minnsta kosti hafa höfundi verið kunnar hliðstæðar samkomur mennskra manna the author at the very least has been familiar with comparable human gatherings 186. This scene is, as Aðalheiður Guðmundsdóttir 2006, 38; 2010, 135 has pointed out, both the earliest mention of hringbrot (a type of ring dance) and the earliest mention of dancing elves in an Icelandic source. As such it can provide us with important information on the history of dance in Iceland as well as the ways in which this popular activity came to be associated with supernatural beings under waves of church-led repression. The refrains sung have been used to garner knowledge on the early forms of popular entertainment in Iceland, and we can see in this a proof of how troll society is frequently used in saga literature as a weird mirror to its human counterpart.

A comparison of this scene in the two versions of the saga shows further how this motif of the troll-assembly can be exploited in different ways. There are a number of additional distinctive details in the longer version. There we are told that yfer dalnum j loptinu komu fram allra handa skip. sumer baurðuzt. enn sumer frỏmdu kaupskap. sumstadar genguzt at fylkingar. ok drap huerr annann. enn blodit rann nidur j dalinn above the valley in the sky all kinds of ships appeared. Some fought, some were engaged in buying and selling. In places troops clashed and killed each other, and the blood rained down on the valley Loth 1962–1965, 52. Such extraordinary elements as flying ships are yet another reason for us to doubt that the shorter version is an abbreviation of the longer: would a scribe really remove such colourful imagery? Whatever the answer, the name of the meeting in the longer version is also different, being referred to by the word Paðreinsleikur. This term (or the variant paðreimsleikur) appears to be derived from the Greek word ἱππόδρομος (hippodromos), and refers to a large gathering involving all kinds of games and sports.The word is probably not a direct borrowing from Greek. Rather it may have arrived in the North via Middle High German poderâm, as implied in Oscar Brennerʼs German translation of Sophus Buggeʼs , entitled 177, and Höfler 283. Descriptions of such festivities appear on more than one occasion in konungasögur when a character makes a journey to Constantinople. Perhaps the most important description is that in chapter 12 of ( The Journey of King Sigurður to Constantinople) in III:281–83, where Sigurður is offered a choice by Emperor Kirjalax between a substantial sum of money or the games, and the wise hero picks the latter.A similar choice is offered to King Eiríkur the Good (hinn góði) according to 237. Applying this name to the games in the longer version of elevates the proceedings somewhat: the troll-party is not just emulating human society now, but the highest echelons of exotic human society. Ríkilát may be the fairest maid in Constantinople, but Þorbjörg rules supreme in her quasi-Constantinople in the mountains.

In the mid-sixteenth-century text of found in Perg. 4to nr 22 in the Kungliga Biblioteket in Stockholm reference is made to the heroes of our saga: Hermann hljóp fyr múrinn mest, mjög var Jarlmann leikinn vest Hermann jumped over the greatest of walls; Jarlmann was really made to suffer Cederschiöld 1883, 65. This condensation of the saga is something of a challenge to those who would rename it Þórbjargar saga digru (see Möbius 117, and the discussion below), the emphasis being firmly placed on the feats and struggles of the two male protagonists.

The vicissitudes of two male protagonists, not just in the world of adventures but also in terms of their personal relationship, is a theme pervading not only the internal dynamics of the saga but also the superstructure into which the saga can be slotted. At the end of the saga we learn that Hermann’s son was Konráður, the protagonist of , and that Jarlmann’s son was Roðbert, the antagonist in the same saga as well as in (where he ends up being defeated by a new set of sworn brothers, Þjalar-Jón and Eiríkur) (see Lavender 2015, 77). It is no mistake that these two additional connected sagas often bear alternative titles emphasizing the double male roles: is sometimes called and also sometimes bears the title .Note that while I use the two-pronged title (following both Rydberg and Loth, but not the inverse form as given by Kalinke and Mitchell), there is variation in the title of the saga as it appears in the manuscript witnesses. Of the three pre-1600 witnesses, none include a title, although AM 556 b 4to has the name Jarlmann written in the margin alongside the first line. AM 167 fol. puts Jarlmann before Hermann–Sögu af Jarlmanni og Hermann kongi–and, as we saw above, Árni Magnússon put Jarlmann first in his notes in AM 576 b 4to. A postscript to in a manuscript containing all three texts from the early twentieth century tentatively proclaims that þessar 3 sögur eru því eiginlega ein sögu að vissu leiti these three sagas are thus in some way a single saga Lbs 2497 8vo, 65. Even if this is the case, few commentators seem to think that they were written concurrently. Bibire implies, for example, that , while dealing with an earlier generation, was actually written after from an opposed view, as if to provide a commentary upon the first saga 70. Kalinke seems to be in agreement, stating that was intended as a counterpart to (and thus must be subsequent to it) as is suggested by the inversion of the fidelity motif in the former 1990, 178.

It is an oversimplification, however, to say that presents a high-born man placing too much trust in an unfaithful companion while presents a high-born man placing too little trust in a faithful companion. itself presents a double plot hinging on the success that can be expected when one trusts a faithful servant and the failure that can result from misplaced suspicion. As Marianne Kalinke has also noted: The plot of reveals a remarkable symmetry. Basic to the structure of the saga is the principle of doubling, expressed through parallelism and antithesis. Instead of one quest for Ríkilát, there are two, the first successfully concluded by Jarlmann; the second, indirectly precipitated by Hermann’s lack of trust. 1990, 176 As we look more closely at , however, and particularly with regard to the shorter version, even these apparently reasonable statements become complicated and further nuanced. For example, alongside the principal doubling of the plot there are numerous other examples of repetition.

There are two quests for Ríkilát, carried out by Jarlmann in the first and second halves of the saga, but in the first half of the saga there are also two battles that to a large extent mirror each other. Jarlmann must defend Ríkilát in combat from the unwanted advances of Rómanus, while Hermann must defend Herborg in combat from the unwanted advances of Ermanus. As already mentioned, in the longer version the names of the opponents are inverted. This adds to the sense that these two battles are more or less interchangeable and consequently not designed to refine our perception of distinct character-traits in our sworn brothers but rather to allow us to see them as perfectly matched and in certain ways indistinguishable. This establishes an equivalence between the two protagonists rather than a contrast.

There are further parallels between Hermann and Jarlmann’s behaviour, however, which function more to distinguish them. For example, Jarlmann (under the pseudonym of Austvestan), on arriving at Rúdent’s court, holds a lavish feast endearing all the people to him. Hermann (under the pseudonym of Norðsunnan) does likewise but to greater effect: Hann gerði og veizlur, gaf gull og silfur vel tvenn slík sem hans fóstbróðir hafði áður gefið, og af þessu varð hann vinsæll og báðir þeir af öllum mönnum he also arranged feasts and gave over twice as much gold and silver as his sworn brother had given before, and because of this he became popular, as both of them were, among all the people. If Hermann reveals himself to surpass Jarlmann in munificence (or self-promotion), Jarlmann shows himself capable of aping his sworn brother’s less appealing characteristics. Just as Hermann displays his jealousy through a sullen silence, so Jarlmann displays problems in affairs of the heart after returning from Kaldbak by becoming mysteriously glum. Pining and showing visible signs of lovesickness is clearly part of the repertoire of both men. The difference in this case, however, is that Hermann’s feelings seem to be genuine (if misguided) whereas context would lead us to assume that Jarlmann’s for Þorbjörg are feigned.

This may not be the whole picture, however. At the start of chapter 23 Jarlmann arrives at the bridal chamber where Þorbjörg is waiting for him. We are told that Hann sest á hennar sængurstokk, en svo hefir hann frá sagt, að honum hefði þá helzt gefið á að líta og efast í, hvort hann skyldi niður leggjast eða ei He sits down on the edge of her bed, and as he tells it, at that point he had the best opportunity to look upon her and doubted whether he should lie down or not.Gillian Fellows Jensen paraphrases this as he is little tempted to lie down beside her Loth 1962–1965, III:62. This statement is somewhat ambiguous: the most obvious reading would be that he is required to lie down as part of the act (leading up to the point at which he will kill this monstrous bride), but the terrifying sight of her makes him hesitate. A minor problem with this interpretation, however, is that the text never actually states that he does lie down, just that he is undressed and then turns to Þorbjörg with the sword and runs her through. There is thus room for an alternative interpretation, that being that perhaps Jarlmann hesitates because he knows that he should kill this woman, but is tempted to actually lie with his bride. We know that she is prodigious in size, but it is not stated that she is ugly as such. Rather we are told in chapter 17 that she is nokkuð stór og hæversk. Aldrei sá hann slíka fyrr, því að hennar hæð tók langt upp hjá fjallinu large and noble-looking. He had never seen anyone like her, because she reached a fair way up the mountainside. Her size is emphasized when she gets stuck in a doorway on another occasion, but there are examples elsewhere that suggest that large women could appeal sexually to male characters (compare Örvar-Oddur’s relationship with the giant’s daughter, Hildigunnur). Her potentially crude verses might, moreover, suggest that she is sexually knowledgeable, and this might add to her appeal. Thus it is just possible that sex with Þorbjörg is not just a test of strength and stamina, but also a test of restraint. Jarlmann must give up the chance of a wild night with an exceptional woman in order to help Hermann out again. In this sense it would be a parallel to his self-control when faced with another amazing (albeit in very different ways) woman, namely Ríkilát. However we interpret these words, it is worth mentioning that this is stated to be Jarlmann’s own account (svo hefir hann frá sagt) of how things happened, hinting that we cannot necessarily trust him to be a reliable narrator.An interesting parallel, which I am grateful to an anonymous reviewer for reminding me of, involves Grettir Ásmundarson. In chapter 35 of the saga that bears his name, after seeing the moonlight shine in Glámur’s eyes, Grettir is said to have sagt sjálfr, at þá eina sýn hafi hann sét svá, at honum brygði við said himself that that was the only sight that he had seen that really shook him , 121. If this is alluded to in Jarlmann’s words, then the intention is not so much to question the reliability of the narrator as to draw attention to the individual’s very personal realization of a moment in which a great deal was at stake and the outcome could have dire consequences (and for Grettir it does).

On yet another occasion we see an example of how Jarlmann’s way with words can confuse situations that appear to be straightforward. Before first meeting Rúdent, Jarlmann instructs his men to let him do all the talking. From this point on Jarlmann’s various speech acts set up the other great doubling within the saga: Jarlmann as Rúdent’s sworn companion as opposed to Jarlmann as Hermann’s sworn companion. At the start of the saga, the whole plot is put in motion when Jarlmann shows an unexpected frankness in reminding Hermann that as great as he may be, he still lacks a queen. The adventures at the court of Rúdent likewise are set in motion by an example of Jarlmann’s direct honesty: Og af því að mér líkar hér vel, þá vildi ég, að engi maður hlyti hér illt af mér. Þó má ei illt varast, nema viti, og vil ég segja yður löst á mér: Ég er svo forvitinn maður, að ég vildi allt vita, en ef ég verð nokkurs vís, þá má ég öngu leyna, og verð ég allt að segja. Ég stend upp um nætur, og geng ég að forvitnast, hvað menn tala, ef ég kann nokkurs vís verða. Því má hver maður varast að tala ekki fleira, en hirða ei, hve nær upp kemur. And because I like it here, I do not want to do wrong by anyone. However, precaution is hardest in ignorance, and I want to admit to you one of my faults. I am so curious that I want to know everything, and if I come to learn of something, I am not capable of concealing it and will say everything. I get up in the night and walk about in order to pry into people’s conversations and see whether I can learn something. For that reason every man must take care not to say more than he cares to come up in discussion. Of course the core honesty is somewhat duplicitous. He has larded his confession with a fair amount of flattery, to soften the truth that he will go around eavesdropping. But he does not tell the truth about why, namely that he is not really Austvestan and that he hopes to learn of Ríkilát’s whereabouts. In the longer version of the saga Jarlmann’s eloquence is even more ambivalent. On learning of Ríkilát from Rúdent he is made to swear an oath. His response is to swear to suikia alldri fruna. enn ek ok aller miner menn skulu þer þiona. suo leingi sem þu lifer nema ek deyia fyrr never betray the woman, and I and all my men shall serve you as long as you live, unless I die first Loth 1962–65, III:48. Since his aim is to have Rúdent killed and save Ríkilát, what appears to be a profession of loyalty is in fact the complete opposite.

Even though this oath is not present in the shorter version, Jarlmann’s undermining of Rúdent is an aspect of the plot that throws into relief the whole system of honour and loyalty that critics have claimed the narrative espouses. If Jarlmann is an example of the perfect servant who should be trusted (as opposed to the duplicitous Roðbert in ), then how should we look upon his seasoned dissimulation at Rúdent’s court? One could rightly argue that, as a pagan king who has abducted a woman engaged-to-be-married, Rúdent has forfeited his right to any faithful service. But beyond his original sin, Rúdent shows no signs of pagan savagery. In fact, Rúdent repeatedly shows a kindness to Jarlmann above and beyond what could be expected. When Jarlmann badgers him to introduce him to his fiancé, Rúdent says ‘Fyrir okkarn vinskap mun eg þetta veita þérʼ ‘On account of our friendship I will grant you thisʼ. When he sees Jarlmann sad, his response is as follows: ‘Góði vinur,’ segir hann, ‘hvað er þér að angri eða ógleði? Eða þykir yður nokkrir hlutir að við oss eða vora menn? Eða eru þeir nokkrir hlutir, að ég má svo gera yður vel líki?’ ‘Good friend,’ he says, ‘what is it that grieves you or makes you so unhappy? Is it something that we or our men have done? Or is there something I could do that would please you?’. Moreover, Rúdent is willing to put his own welfare on the line for Jarlmann, as when he negotiates a match: ‘En fyrir okkarn vinskap mun ég við leita, að þitt mál takist, hvernig sem mér veitirʼ ‘But for the sake of our friendship I will try to ensure that your cause is successful, however it turns out for meʼ. There are numerous other small examples that make Jarlmann’s ultimate betrayal of Rúdent seem somewhat awkward, and it is hard to draw a straight moral conclusion. Perhaps the point is that no matter how kind someone is to you, sworn-brotherhood wins out (even when your sworn brother is surly and distrustful). Perhaps too we are forced to accept the for-the-time uncontroversial maxim that an obligation to a Christian always trumps that to a pagan. Yet alongside this, the saga may also be probing the grey areas of sworn-brotherhood, showing how the demands it imposes force one to neglect one’s own welfare and other types of meaningful relationships. However we read the saga, it is once again clear that in the focus on the relationships between Jarlmann and his two patrons, homosocial interactions (at times mediated via the female characters) are at the core of the ethical structure presented.

One final example where a strange kind of anxiety arises by following the narrative logic through to its natural conclusion is in the scene of the ring. With Hermann being such a fine specimen of a king, the best in all the world, it seems strange that he needs a magic ring to woo the equally excellent daughter of the emperor.If he is not a fine king, then we would need to question Jarlmann’s loyalty (since Jarlmann claims that the only thing that Hermann lacks is a wife). Surely we should not see Jarlmann as a false hypocrite praising a less-than sterling monarch? One might imagine that, all obstacles being absent (such as a curse, for example, present in many similar bridal quests, but not relied upon here), Ríkilát would choose voluntarily to be with Hermann. But in this case we may be concentrating too much upon the consent and/or coercion of the female character. It seems possible that the introduction of the ring was carried out without any consideration of female agency, rather as a way of emphasizing the nature of the masculine bonds. Jarlmann is a proxy wooer in possession of an object that can guarantee that an individual will fall in love with whomsoever the bestower desires. This surely represents an extreme temptation, as he could use it to ensure that Ríkilát falls in love with him instead of Hermann (it may even be implied that this would be a more fitting match). The fact that Jarlmann does not exploit this possibility is, for good or bad, proof of his fidelity. Bagerius 113 says that is one of the works that shows that Tristrams moraliska dilemma fick återverkan på Island Tristanʼs moral dilemma (i.e. that between oneʼs lord and oneʼs beloved) had repercussions in Iceland, even if the outcome is different: Jarlmann places Hermann above Ríkilát. Even with a foolproof method to usurp the place of his sworn brother in Ríkilát’s affections, Jarlmann holds back and does the honourable thing. That Ríkilát now has no say in the matter would seem to be considered beside the point.

It is also possible to see the episode involving the ring from a slightly different angle that accepts Ríkilát as a more autonomous character. Perhaps the ring is necessary precisely because Ríkilát has amorous feelings for Jarlmann. It thus acts not so much to ensure her love for Hermann but to eclipse her love for Jarlmann. A suggestion of potential feelings on Ríkilátʼs side could be interpreted when Jarlmann, in the shorter version, draws an image of Hermann, and Ríkilát responds Víst ertu mikill meistari, og ekki kan ég ætla, að nokkur mundi sig öðruvís kjósa, þó sjálfur ætti um að ráða You are clearly a great master, and I cannot imagine that anyone would want to look any different, even if they were free to choose. A straightforward reading would see the first part as referring to Jarlmann and the second part as referring to Hermann (based on the image produced), but it is just as possible to imagine both parts of this phrase as directed towards Jarlmann (i.e. having seen Jarlmann alongside the image she still cannot see why anyone would want to look any differently than Jarlmann does). Jarlmannʼs response that she would feel more strongly if she met Hermann in person does not clear up the ambiguity.It is interesting that Sävborg (564), talking of the indigenous romances, says that kärleken är i många fall ett styrande drag för figurerna själva love is in many cases a driving trait of the characters themselves and i många fall framstår det som en okontrollerbar kraft som styr figurerna och deras uppträdande. Det gäller … Ríkilát í in many cases it (i.e. love) appears as an uncontrollable force that governs characters and their actions. Such is the case for … Ríkilát in . The idea of Ríkilátʼs strong love would seem to be based principally upon this (ambiguous) statement of hers and then upon her statement after being charmed with the ring that she cannot think of anyone but Hermann. An oral reading of the saga, however, might have presented opportunities to push the reading one way or the other, through the use of suggestive gestures or tone. Likewise, after the suspicious Hermann dismisses Jarlmann from his court, we hear that Ríkilát var jafnan ókát, síðan Jarlmann fór á brott, og lét þó sem minnst á sér finna Ríkilát was always unhappy after Jarlmann had left, but did not let it show at all.

It may be that Jarlmann loves Ríkilát too. Hermann certainly seems to think so, and when faced with the latterʼs jealousy, Jarlmannʼs statement that he did not want to betray his lord does nothing to dispel the implication that he has a strong affection for his lordʼs betrothed. The capacity of the saga for psychological sophistication is emphasized by its recognition of the possibility of psychosomatic illness: when Jarlmann, seeking to engineer his first audience with Ríkilát, feigns sickness, Ríkilát vill nú skoða og rannsaka hans sjúkdóm, hvernig honum sé háttað. Síðan talar hún til hans: “Ég hefi hugsað um hríð, og finn ég ei sótt í líkama þínum, nema svo sé henni háttað, sem orðskviðurinn hljóðar, að ‘sá er ekki heill, sem hugurinn veikir’, og hygg ég, að svo muni vera.” now wants to observe and investigate his illness and how it presents itself. Then she speaks to him: “I have thought for a while, and I cannot find any illness in your body, unless it is the kind of thing where, as the saying goes, ‘he is not whole, whom the mind weakens.’ And I think that this is the case here.” Indeed, Jarlmann apparently feigns emotional problems on two occasions: the one described above and then when he feigns lovesickness in order to secure Rúdent’s aid in arranging his marriage to Þorbjörg. The success of his feigned ennui after meeting Þorbjörg suggests a familiarity with genuine lovesickness. It may even be bona fide lovesickness—the true object being Ríkilát, whom he has seen encaged by Þorbjörg—which he encourages to be misinterpreted as evoked by Þorbjörg. If this is the case, then maybe when Jarlmann is hugsjúkur sick at heart, dejected in Constantinople, it too is not merely an act but genuine longing for Ríkilát. And Ríkilát, skilled as she is, may well be aware of the true nature of Jarlmannʼs illness. Early on she asks him to teach her Frakklands vísu the ways of France, which may be a subtle euphemism for the arts of love, and her father certainly sees the pair as a promising match. Of course, even if we accept this mutual affection, it is still clear that both Jarlmann and Ríkilát make the difficult decision to submit themselves to other demands, namely duty to a lord and obligation to marry a social equal. This sacrifice involves them sublimating their feelings for each other and experiencing it only vicariously through the conduit that is Hermann. It is worth remembering also that the ring that ensures Ríkilátʼs dedication to Hermann is given to Jarlmann by Herborg, Hermannʼs sister. It is natural that she should wish her brotherʼs suit to end well, but it may be that even at this early stage of the narrative she knows that Jarlmann is who she is destined to end up with. Giving him the ring not only helps her brother, but it also puts Jarlmann to the test. As she emphasizes, varðar þá miklu, hversu sá vill venda, er með fer what matters the most is how he who has it wants things to turn out. Jarlmann passes the test, stays loyal to his lord and eventually marries Herborg. While they are said to fall in love, Jarlmann eventually choses to end his life in quiet contemplation, and this too may be a subtle hint at an unfulfilled alternative love, as many romance characters who have been frustrated in their amorous pursuits chose religious seclusion at the end of their lives.

, like many of its generic bedfellows, was more or less ignored by scholars of Old Norse literature until the start of the nineteenth century. Prior to that time it was mentioned in George Hickes’ 1703–1705 II:314 as well as the 1777 of Hálfdan Einarsson (1732–1785) under the category of historiæ veterum, qvæ nationes exteras spectant stories of ancient peoples that concern countries outside of Scandinavia 100 as Jallmanni and Hermanni, Gallorum of the Frenchmen Jarlmann and Hermann 103. In the same year Hermans og Jarlmans s. appears in a list of Icelandic literature found in 1777, 157 by Uno von Troil (1746–1803), a work based on the Swedish bishop’s experiences while visiting Iceland in 1772 along with the British traveller and collector Sir Joseph Banks (1743–1820). The saga is also listed in Peter Erasmus Müllerʼs 1819–1820, III:482 along with other oversatte eller efterlignede romaner translated romances or imitations of romances III:480.

At the start of the nineteenth century a translation appeared for the first time, as the third offering in Johan G. Liljegren’s (1791–1837) 1818–1819, a work that claims in a subtitle to be till läsning för Sveriges ungdom intended as reading for the youth of Sweden. The previous two sagas to be served up in translation were and , both accompanied by extensive notes. Both Hrólfur and Oddur are recorded in their narratives as having been born in regions of modern-day Norway (although their adventures take them much further afield) and so have a good claim to be called heroes of the Scandinavian legendary age. Jarlmann and Hermann, conversely, seem hardly worthy of the bestowal of such a title. Nevertheless, Liljegren showed a sensitivity to the relevance of such romance sagas—particularly their potential to lure new converts to saga literature—which can lead us to see him as in some respects ahead of his time. Whereas others had simply ignored the vast swath of romance literature dealing with more southerly Europe and lands surrounding and beyond the Mediterranean, in his introduction Liljegren states (albeit in a footnote) that the saga is ur en Sagokrets, hvoraf Öfvers. velat meddela endast denna lilla Saga såsom prof af den der rådande andan from a group of sagas of which the translator wanted to share just this little saga as a proof of the prevailing spirit of the time xix.

The introduction reveals that, like many, Liljegren was fascinated with exotic escapades of Northern adventurers in the East on the route to Constantinople. Örvar-Oddur makes it to Aquitania and as far as the River Jordan in his saga , II:267–71, but what really interests Liljegren is the conflict between Oddur and Kvillánus blesi (or Ögmundur Eyþjófsbani in disguise), which takes place in parts of modern-day Russia and Ukraine between the Baltic and Black Seas. Liljegren follows Peter Frederik Suhm in seeing this conflict as a confused retelling of that between Oleg of Novgorod (ruled 879–912) and the Greek emperor Leo the Philosopher (ruled 886–912). It is this interest in the eastern route that can help explain his inclusion of . Here we have royal personages whose dominions include Garðaríki (see below) and who send envoys back and forth to Constantinople. For Liljegren, Hermann’s wooing mission seems to be part of the same route of cultural contact as that for which Örvar-Oddur’s battles are a cipher.

Liljegren was working with manuscripts found at the Kungliga Bibliotek in Stockholm,A reference to Niflunga saga where all of the other 70+ manuscripts of have Hjaðninga saga, for example, shows that either Papp. 4to nr 17 or Papp. 8vo nr 8, both manuscripts in the Kungliga Biblioteket, was Liljegren’s source for the translation of that saga. and at the time he was active two copies of were found among the holdings there, in Papp. fol. nr 56 and Papp. 8vo nr 4. The text in the latter bears the title Sagan af Þorbyørgu dygru f. 29r, suggesting that it was not Liljegren’s primary reference (in his introduction he refers to som i en annan handskrift kallas Thorborg digras saga which in another manuscript is called the saga of Þorbjörg the stout xix; my italics. A quick comparison of Papp. fol. nr 56 and Liljegren’s translation strongly suggest that the former was his primary source. Gödel 167 explains that the manuscript was written by the Icelander Arngrímur Jónsson at the Antikvitetskollegium in the period during which he was employed there, that is to say between 1683 and 1691.This Arngrímur, whose dates of birth and death are uncertain (see also Busch 23), is not to be mistaken for the more famous Arngrímur Jónsson (1568–1648), author of, among other works, , and who was active about a century earlier. Arngrímur has included brief chapter headings (not present in other manuscripts) that correspond almost identically to the chapter headings in the translation (e.g. chapter 2 on p. 4 of the translation corresponds to on f. 2r of the manuscript). The first sentence of the translation also exposes this relationship: Wilhelm is said to rule over Gardarike, Tyskland, and parts of Frakkland, as too in Papp. fol. nr 56 (yet not in either the longer or the shorter versions as presented by Rydberg or Loth, where no more than Frakkland is mentioned).The text in Papp. fol. nr 56 is, however, clearly closer to the shorter version than the longer version. Another manuscript at the Kungliga Bibliotek, Engeström B: 3. 1,21, confirms these suppositions. It contains Liljegren’s personal transcription of the saga, alongside which he has included variants, principally from a manuscript that he calls Þ (i.e. Papp. 8vo nr 4to, Þorbjörgs saga digra), but also at times from a manuscript that he calls J (Papp. fol. nr 56, noted as a variant on the rare occasions when he preferred a reading from the former manuscript).

Papp. fol. nr 56, moreover, is a perfect example of a manuscript that could have had an impact upon Liljegren’s understanding of the saga. It contains thirteen texts, seven of which are fornaldarsögur, four of which are riddarasögur, and two of which are Íslendinga þættir.The full contents are as follows: 1. Jarlmanns saga og Hermanns; 2. Orms þáttur Stórólfssonar; 3. Þorsteins þáttur bæjarmagns;
 4. Hálfdanar saga Brönufóstra;
 5. Egils saga einhenda og Ásmundar berserkjabana; 6. Samsons saga fagra; 7. Blómstrvalla saga;
 8. Fríðþjófs saga frækna;
 9. Sigurðar saga þögla;
 10. Hróa þáttur heimska;
 11. Illuga saga Gríðarfóstra; 12. Sturlaugs saga starfsama;
 13. Sörla saga sterka. The more rigid generic divisions that exist in our period were still under negotiation at the start of the nineteenth century, and prior to Carl Christian Rafn’s corpus-defining multiple-text edition (1829–1830), one could be forgiven for not seeing any significant difference between the content and themes of texts such as and and those of .

The notes that Liljegren appended to his translation are not as copious as those accompanying the previous two sagas. Besides brief comments on the practice of giving rings, monetary systems, galleons, and portraiture, there is one long excursus on medicine and healing in ancient Norse culture. This is in response to the mention of the cross-mark on Ríkilát’s hand that conferred healing powers, but the discussion goes way beyond its humble origins to engage with various claims. Examples are the idea, put forward, that most vikings only suffered from illnesses similar to the common cold; the suggestion that the emphasis on fate and bravery in ancient Scandinavian culture meant that medicines in general were not highly esteemed; and the argument that the human need for explanations led Liljegren’s ancestors to attribute causality to spirits and supernatural powers where the curative process was not fully comprehended. Some of Liljegren’s conclusions may be dubious, but the way he presents them can certainly be seen as entertaining.

Unfortunately Liljegren’s contemporaries did not quite see things this way. In the we read that: På egen begäran fick han 1817 till bl a Tegnérs förargelse i uppdrag att översätta o utge på KB förvarade isländska manuskript. Han valde fornaldarsagorna, men resultatet, som trycktes under titeln Skandinaviska fornålderns hjeltesagor, blev inte någon förlagsframgång. Av planerade tio band utkom två o de var alltför lärdomstyngda för att nå en större publik. Arbetet hade dock betydelse för att sprida kunskap om de isländska sagorna i Sverige på 1800-talet. Broomé 18 At his own request and to the annoyance of, amongst others, Tegnér, in 1817 he was commissioned to translate and publish Icelandic manuscripts kept at the Kungliga Bibliotek. He chose fornaldarsögur, but the result, which appeared under the title , did not end up being a great success for the publishers. Of a planned ten volumes, two came out, and they were much too weighed down with learned content to reach a wider public. The work was, however, significant in disseminating knowledge about the Icelandic sagas in eighteenth-century Sweden. Thus marked the end of the series, and the series marked the last appearance of the saga in print until Hugo Rydberg’s edition of the shorter version in 1917, just under a century later. This translation carries on where Rydberg left off and, like Liljegrenʼs translation, has been produced in the hope of making the saga accessible to a new generation of scholars, both young and old.

Chapter numbers are not consistent throughout the manuscript transmission. The chapter numbers provided here are based on Rydbergʼs and Lothʼs editions and are intended to facilitate comparison. Rydberg bases his chapter divisions on AM 529 4to and on AM 167 fol. where the former witness is defective. Neither of these manuscripts contain numbered chapters, but AM 529 4to clearly signals textual divisions by leaving an indented space, normally of two linesʼ height, where an initial could be inserted (although they ended up not being added). AM 167 fol. uses slightly increased spacing and larger initials to signal textual divisions. Loth based her chapter divisions on AM 556 b 4to, which, like AM 529 4to, has no numbered sections but signals textual divisions by leaving a blank indented space of between two- and three-linesʼ height where an initial could be inserted. Note that in the table below all significant differences between the two versions are highlighted in italics.