¹ See an article by the present writer, Disputed questions in Beowulf-criticism, in the Publications of the Modern Language Association of America, vol. xxiv (1909), pp. 220 ff., esp. pp. 237 ff. These relationships have been recently discussed in a dissertation by O. L. Olson, The Relation of the Hrólfs saga kraka and the Bjarkarímur to Beowulf, in Publications of the Society for the Advancement of Scandinavian Study, Urbana, Illinois, 1916. I do not find my conclusions altered by a reading of this dissertation, and I think it impossible, as stated in the article above, to reach definite results with such scanty and elusive evidence. When Dr. Olson has worked with these problems a little longer I think he may become less dogmatic.

The resemblances between the fight with the dragon in Beowulf and Frotho's dragon-fight in Saxo, pointed out by Sievers, appear to be commonplaces of dragon story, and not proof of a common origin of the two episodes. This position, has been stated at length and convincingly by Olrik, Danmarks Heltedigtning, Copenhagen, 1903, vol. i, pp. 307–315. Those who do not read Danish easily will be interested in a forthcoming translation of the Heltedigtning, to be issued by the American-Scandinavian Foundation, New York City.

² F. Panzer, Beowulf, Munich, 1910; pp. 296 ff.; esp. p. 305. The generalizations of this author must be treated with some reserve; he is better as collector than as critic.

³ Edward Topsell's History of Serpents, London, 1608, etc., abounds in quaint quasi-scientific observations. Topsell classified dragons after the fashion of a Buffon or a Cuvier: Class I has wings and no feet; Class II has 'both wings and feet; Class III has neither wings nor feet; and so forth. Illustrations from Topsell are reproduced in Shakespeare's England, London, 1916, vol. I, p. 497.

⁴ The text of Wyatt and Chambers, Cambridge, 1914, has been used. The reader should also consult Dr. Chambers's forthcoming Introduction to the Study of Beowulf.

⁵ The word meldan is ambiguous; it may mean either “informer” or “finder.”

⁶ Cambridge History of English Literature, vol. i, p. 26.

⁷ Beowulf, Manchester, 1913, p. xxv.

⁸ Beowulf, Heidelberg, 1906, p. 116.

⁹ Clark Hall, Beowulf, London, 1901, p. 113. See also Bugge, Paul and Braune, Beiträge, vol. xii, pp. 370–1.

¹⁰ There are other choices than these two words for the missing letters. We might read þreoe or þrym, for instance, and translate “but for dire need fled the violence of some one of the children of men, his blows of hate.” It will be noticed that there is no evidence as to the rest of the word in the ms. as it stands at present. “The last three letters are now quite illegible” (Chambers). Kemble read þegn, and Zupitza thought that the traces of letters justified that reading. According to the Zupitza autotype, there does not appear to be room for a word of more than four letters.

¹¹ As Schücking notes, mid gewealdum (2221) is parallel to sylfes willan, and similar in meaning (Vocabulary, ed. of 1910).

¹² Consult Dania, vol. i, pp. 233 ff.: “R⊘veren ved Gråsten og Beowulf,” I af Sophus Bugge; II af Axel Olrik. Panzer's discussion will be found on pp. 308 ff. of his Beowulf, Munich, 1910 (see p. 549 above).

The Danish popular tale cited by Bugge is short; it may be given entire in translation.

“Once upon a time there was a robbers' cave near Greystone. In it dwelt twelve thieves, who employed the device of stretching a cord across the road, so that when travellers came by, a bell rang in the cave. But since they were well concealed, it happened that one after another of them died [a natural death], so that finally only the twelfth remained. He was very old and had a long grey beard. At the very end of his life, he was walking through the woods, when he met a man to whom he promised to give a big chest, full of gold and many precious things, provided he would bury him when his time came. But the robber stipulated that the chest should not be opened nor anything taken out, before the man got across the water.

“It happened to be winter-time when the old robber died and the man buried him, so that the chest had to be hauled over the ice. And, as is customary, the men who were dragging the treasure were cautioned to be silent and not to speak a word, until they got the chest to land. But as they were pulling hardest, one of those who were drawing the chest forgot to hold his tongue, and immediately the treasure sank down through the ice. Now one can only feel the chest, when one searches for it with a pole at that place.”

Bugge remarked that if the connection which he believed he had established between the passage in Beowulf and this tale were correct, “the explanation may be that this episode, like so many others in Beowulf, rests on an old Danish narrative [en dansk oldsagn] (p. 235). Olrik finds that the resemblances pointed out by Bugge are matters of detail, and that the essentials of the two narratives are not in agreement,—much the same sort of argument, it will be observed, as he used in refuting Sievers' parallel between the dragon in Beowulf and the Frotho dragon in Saxo. Nor does Olrik believe the ”R⊘veren ved Gråsten“ necessarily the development of an ancient traditional tale; the hero is elsewhere unknown, the nameless champion in Beowulf does not prove his existence, and a traditional tale preserving so many details in modern Danish, would have been more individual [enestående]. (241)

¹³ Stjerna (see note to p. 579 below) took account of both these narratives, and perceived their discrepancies. He regarded the elegiac passage, 2231 ff., as late in its present form, because “a collection of such magnitude as is described in Beowulf, and consisting of objects of the kind specified in the poem, can evidently not have been concealed by a solitary person” (p. 143), thus confusing poetic imagination with archaeological fact. Although his argument is wrong, his conclusion appears to be right, for other reasons than he perceived.

¹⁴ Since completing this article, I notice that the emendation næfne for næs has already been suggested by Holthausen, in the 4th edition of his Beowulf, Part i, p. 99 (Heidelberg and New York, 1914). I have been unable to procure a copy of Part II of the same edition, and the New York publishers inform me that it has not yet been issued. Consequently I am unable to say how far the rendering suggested above represents Holthausen's understanding of the passage. I imagine, however, from Holthausen's reading goldhwæte[s] that he does not take āgendes to mean God', which gives a very different sense from the reading here proposed.

¹⁵ The meaning of the phrase gold-hwæt is difficult to render in modern English. It combines the significance of “active, keen, bold” (See Chambers, Glossary, sub hwæt) with the idea of the possession of gold. Perhaps the word “prosperous” might come near the meaning: I have rendered it here “rich in gold” in order to keep the double significance of the phrase. It is quite in place as applying to the secg of 3071, who might plunder the hoard. Cf. fyrd-hwæt (1641 and 2476), applied to the Geatas who accompanied Beowulf to Heorot, and to the sons of Ongentheow. The rendering “greedy for gold,” favored by some editors, seems to me to read into the phrase a (meaning which does not belong to it, although it might assist the interpretation of 3074–5 suggested above, as pointing with added sharpness the antithesis 'between the man's possessions and his piety. The ms. reading gold-hwæte may be retained, if preferred, and the word understood adverbially, as by Holthausen in his third edition. It seems to me more in accord with Anglo-Saxon idiom to make it an adjective.

¹⁶ Publications of the Modern Language Association of America, vol. xxvii (1912), pp. 222 ff.

¹⁷ See below, p. 570.

¹⁸ See Panzer, loc. cit., p. 296.

¹⁹ Cf. Chambers and Sidgwick, Early English Lyrics, London, 1907, p. 282.

²⁰ Dania, vol. i, p. 271.

²¹ See Bosworth-Toller, under

²² The seems properly to denote the grave-chamber of stone within (cf. under hrūsan, 2411), but is often used loosely of the whole structure (cf. 3169). There is an elaborate discussion of the term beorh, , stān-boga, etc., in Schücking's Untersuchungen zur Bedeutungslehre der angelsächsischen Dichtersprache, Heidelberg, 1915. Most of the conclusions seems to me untenable, however. Schücking takes as his point of departure the theories of Stjerna, which I am not at all concerned to defend, but it is a little difficult to separate Schücking's constructive argument from his criticism of the Swedish scholar. Schücking's main thesis is that we are not dealing with a grave-mound at all, but with a natural cave in a hill. He argues that beorh sometimes means a height, not a cairn, which no one will deny, but he does not meet the objection that it sometimes just as clearly means a barrow, as a glance at Bosworth-Toller will show. He admits that it is difficult to reconcile the adjectives eall-gearo and nīwe with a natural hill or mound; that they seem rather to indicate an artificial elevation. So he takes refuge in dogmatic assertion: “nīwe means ‘unused.‘” But what evidence is there for this? He thinks that a stream could not issue from a grave-mound, and that this is further proof that the mound is not artificial. But the stream is not an ordinary brook; it is a part of the supernatural “machinery” of the dragon, as the phrase “hot with battle-fires” shows. We need not examine too closely into these effluvia from the physiological point of view, but if the reader will let his imagination play about the situation a bit, I do not think he will find anything strange about it. Schücking's theory gets him into further difficulties in connection with 2718–19; see below, p. 575, note 31.

Since the main part of this article was written, I have noticed that Holthausen, in the second volume of the third edition of his Beowulf (1913), reproduces two illustrations of Danish burial-mounds from Müller's Vor Oldtid; see also his notes, p. 150.

²³ See Thurman, Archæologia, vol. xlii (1869), p. 202, and note b.

²⁴ See below, p. 577.

²⁵ On this general subject, consult for Sweden, O. Montelius, Civilization of Sweden in Heathen Times, trsl. F. H. Woods, London, 1888; for Denmark, Sophus Müller, Vor Oldtid, Copenhagen, 1897; for France, P. de Mortillet, Origine du Culte des Morts; les Sépultures Préhistoriques, Paris, 1914; for Great Britain, Thurnam as above, and in general L. Reinhardt, Der Mensch zur Eiszeit in Europa, Munich, 1908, and J. Fergusson, Rude Stone Monuments in All Countries; Their Age and Uses, London, 1872. For further bibliography, see Müller, p. 194.

²⁶ “In the great chambered tumulus of Maeshow in Orkney, which, from the Runic inscriptions on its walls, seems to have contained much treasure, there is actually the figure of a dragon drawn with much art and archaic skill” (Thurnam, p. 204).

²⁷ See an article by the present writer in these Publications, vol. xxvii; esp. pp. 240 ff.

²⁸ Usually explained as Breton tól, “table” and men, “stone,” but this etymology is not wholly satisfactory. See the New English Dictionary.

²⁹ Montelius, loc. cit., pp. 33–34.

³⁰ This was recognized by Thurnam: “In the very early Anglo-Saxon poem of Beowulf there is notice of what was evidently a chambered tumulus.” The context shows clearly that Thurnam had the passage tumulus, not the cist-type, in mind (Archæologia, loc. cit., p. 202).

³¹ See illustrations in works cited above, especially the discussion by Thurnam in Archæologia. Schücking has a good deal to say of the phrase stān-boga in connection with his theories about the beorh, commented upon above, p. 570, note 22. In reply to his uneasiness about the construction of 2718–19, one may quote the excellent note by Chambers: “No change is necessary. For the tense cf. ll. 1923, 1928, 2486; and for the sg. verb with plu. subject in a subordinate clause cf. l. 2164, … Further I do not see why should not be the subject: ‘How the earth-hall contained within itself the arches… .’” There is really no difficulty here. But the phrase enta geweorc (2717) is difficult to reconcile with Schücking's theory that beorh is a natural hill, and the lair of the dragon a cave. So he explains enta geweorc as referring, not to the construction of the chamber in which the treasures lie hid, but to the treasures themselves. He does not give a translation of the passage as he conceives it should be read, and I confess it puzzles me to understand how it can be construed as he wishes. As a test of the validity of Schücking's theory, let any one read these three lines, translating enta geweorc as “giants’ gold,” or something of the sort, and see how he comes out.

³² “A central avenue or gallery, having a doorway or entrance at one end, by which it was entered … may indeed be regarded as the essential character of a sepulchral chamber, as distinguished from a vault or cist, of however large proportions, the interior of which can only be reached, after raising the covering-stone from above” (Thurnam, p. 212).

³³ Montelius, loc. cit., p. 32.

³⁴ See discussion of the phrase in the article on the Haunted Mere just referred to, pp. 213 ff.

³⁵ Müller, Vor Oldtid, p. 80. Müller is of course particularly concerned with Danish tombs, hut these seem to have been essentially the same as those in Sweden.

³⁶ For a summary of these differences, see Müller, loc. cit., p. 86. The following description from Thurnam, pp. 212 f., gives a more extended account of British conditions: “The central avenue or gallery is situate at the broad end of the tumulus, and, like the side chambers often opening out from it, is formed of two rows of stones set on edge, supporting others laid horizontally across, and having interstices between filled up with horizontal walling, similar to that described as supporting the base of most of these mounds. In the finest examples of chambered barrows, as those of Uley, Stoney Littleton, and Nempnet, the entrance to the avenue is, or was, by a well-built doorway, formed of two standing and one transverse or horizontal stones, which three stones (trilithon) are, for the most part, of larger and more massive proportions than any of the others entering into the composition of the chambers. This doorway is found several feet within the skirt or general base-line of the tumulus… . The entrance, varying from two and a half to four feet in height, was closed by a large stone on the outside, which could be rolled away as required, and was itself covered with the rubble-stone and earth of which the barrow in general was formed.” An understanding of these conditions will be much aided by an examination of the plates in Thurnam's essay. For conditions in Scandinavia, see especially Müller's illustrations, pp. 72, 74, 75, 79, 82 ff.; Montelius, pp. 30 ff., and Reinhardt, esp. p. 572.

³⁷ See Thurnam, plate to face p. 212,

³⁸ See p. 575 above.

³⁹ See illustrations of the barrow at Uley, Thurnam, p. 213, and cf. the mound at Uby in Zealand, reproduced by Reinhardt, p. 572.

⁴⁰ K. Stjerna, Essays on Questions connected with the old English Poem of Beowulf, translated and cited by J. R. C. Hall, London, 1912, p. 140. Stjerna's long article on the Dragon's Hoard is, I think, wholly astray in its main conclusions, though it provides some interesting descriptions and illustrations of archaeological remains. It does not seem necessary to criticise his work in detail here, but the nature of its errors may be briefly indicated. Some of these are pointed out in the notes provided by Dr. Hall. Stjerna's understanding of Anglo-Saxon is very often at fault, as in the phrases cited above. Another article in the volume, “The Double Burial,” is founded upon such a misapprehension; see Hall's note, p. 171. Stjerna is too much inclined to take statements in the poem literally, forgetting that we cannot treat the decorative passages as if they were a museum catalog, from which their age and period might be determined, and he regards poetic Action as if it gave trustworthy information about the origin of the treasure. Thus he says, p. 143: “A collection of such magnitude as is described in Beowulf, and consisting of objects of the kind specified in the poem, can evidently not have been concealed by a solitary person.” In short, this work, though conscientious and learned, and sometimes very suggestive, is likely to prove misleading unless constantly checked up and corrected. Stjerna sometimes builds arguments upon hypotheses discussed earlier in the book, which he regards as proved.

⁴¹ See above, p. 577.

⁴² Loc. cit., p. 208. It is unfortunate that 2212 cannot afford assistance in deciding these questions; the line is hopelessly obscured in the manuscript. Chambers conjectures sē on hēa[um hope] hord beweotode, and comments as follows: “Sedgefield reads , ‘on the high heath’; but is feminine; was also read by Sievers in 1870–1 [P. B. B., xxxvi, p. 418], so this is probably to be taken as the ms. reading. However to me it looks more like hēaum hope, ‘on the high hollow.’ The word hop survives in Northern English hope, ‘a hollow among the hills,’ as, for example, in Forsyth, Beauties of Scotland: ‘The hills are everywhere intersected by small streams called burns. These, flowing in a deep bed, form glens or hollows, provincially called hopes.’ Although by the sea, the mound may have stood in such a hollow or hope: cf. the mention of the burn in ll. 2545–6.”—This conjecture does not seem to fit the location of the barrow, as our examination of the poem has revealed it. And the stream, “hot with battle-fires,” is surely a kind of supernatural dragonish outflow, not an ordinary brook. The context makes it clear that the beorg out of which it runs is the grave-mound itself, in which the dragon lurks. This is what might be expected if the stream is an outflow of the dragon's superfluous energies; it would be very strange to have a burn of the ordinary sort flowing out of a sepulchral mound.