^{Note 1 in page 939} There are three redactions of the text of the shorter list; (1) in the V$ooluspá in codex regius of the Elder Edda; (2) in slightly different wording, in codex regius of the Younger Edda, Gylfaginning, Chapter 14; (3) in the version of the V$ooluspá found in Hauksbók.

^{Note 2 in page 939} They are in two manuscripts which contain the Younger Edda, A. M. 748, II, 4to, and A. M. 757 4to. Best accessible in Finnur Jónsson, Den norsk-islandske Skjaldedigtningen (København, 1912), A I, 679 f.; B I, 672.

^{Note 3 in page 940} The variants of the dwarf-names in the V$ooluspá have been collected three times; (1) by Sophus Bugge, Norrœn FornkvœÐi (Christiania, 1867), 27 f., (2) by Eugen Mogk, PBB VII (1880), 249 ff., (3) by B. Sijmons in Sijmons and Gering, Die Lieder der Edda (Halle, 1906), I, 20 f.

^{Note 4 in page 940} Otto Jespersen is in error when he says, “In the poetry of the Old Norsemen we again find a liking for meaningless sounds. Take the names for the dwarfs in the Snorra Edda:—” and he quotes V$ooluspá 11 and 12 in the Younger Edda. “Several of these names recall well-known words. Most of them however are certainly empty sounds, but as such of great effect.” He follows this with a list of Othin-names of which he says, there are “no few that are mere empty sounds, unconnected with anything known,” with more to the same effect. The Othin-names had been explained in 1924 by Falk in his Odensheite, cited infra. Jespersen's remarks are in his “Mankind, Nation and Individual from a Linguistic Point of View,” Institutet for sammenlignende Kulturforskning. Serie A: Forelesninger IV (Oslo, 1925), 188 ff. Zachrisson calls attention to Jespersen's error in Studia neophilologica, I (1928), 85.

^{Note 5 in page 940} In a previous article “They who await the second Death,” Scandinavian Studies and Notes IX [1927], 167-201) I have shown in detail the characteristics of dwarves as recorded in the Icelandic Romantic Sagas.

^{Note 6 in page 940} So much time has intervened between the beginning of this study and its completion that I can not indicate in all cases which definitions and etymologies are my own and which are due to others; and so I have abstained from branding any of my property and must assume responsibility for the errors of my predecessors as well as of myself. I am fully aware that many of the definitions and etymologies offered in the following pages are problematical, but I hope that this tabulation will result in others bringing better solutions.

^{Note 7 in page 940} Since the occurrences of each name are ordinarily listed in E. H. Lind, Norskisländska dopnamn och fingerade namn från medeltiden (Upsala, 1905-15), or Finnur Jónsson, Lexicon poeticum antiquœ linguœ septentrionalis (København, 1913-16) or in both, it is necessary to give this information only in cases where for some reason it is not readily found in the above works. The former is here referred to as Lind, the latter as FJ. FJ defines many of these names, as does also Hugo Gering in his Kommentar zu den Liedern der Edda, erste Hälfte (Halle, 1927), referred to as Gering, followed by the page number in parentheses. References to the Elder Edda are based on Gustav Neckel, Die Lieder des codex regius nebst verwandten Denkmälern, (2nd ed., Heidelberg, 1927), references to the þulur on Finnur Jónsson's Skjaldedigtningen. All Modern Icelandic words are found in Sigfús Blöndal, Islenzk-dönsk orÐabók (Reykjavik, 1920-24), all Modern Norwegian words in Alf Torp, Nynorsk etymologisk Ordbok (Kristiania, 1919). Words from other Scandinavian dialects than Old Icelandic are quoted from Torp unless specified. The language is not designated for Old Icelandic words found in Fritzner's Ordbog over det gamle norske Sprog (Kristiania, 1886-1896), in Vigfusson's Icelandic-English Dictionary (Oxford, 1874), or in FJ. I do not stop to disagree with Gering in the many instances in which he wanders far afield for an etymology of a dwarf-name. Gering did not observe that the dwarf-names were made from the living linguistic material used and understood by the þulr and his contemporaries. Instead he often tried to derive dwarf-names from ancient compounded hero-names, the parts of which were so fused that none but a learned etymologizer could separate and understand them, the meanings of which would be a sealed book to the simple people who made up the dwarf-names. But I gratefully acknowledge here my indebtedness to him.

^{Note 8 in page 942} Hermann Güntert, Von der Sprache der Götter und Geister (Halle, 1921), 141.

^{Note 9 in page 942} Gustav Indrebo, “Nokre stadnamn,” Heidersskrift til Marius Hœgstad (Oslo, 1925), 65 ff.

^{Note 10 in page 943} Noreen, Altisländische und altnorwegische Grammatik ⁴ (Halle, 1923), §318, 8.

^{Note 11 in page 943} For mythical connotations see references in Bugge, FornkvœÐi 3a, New English Dictionary s. v. brownie and browning, Hj. Falk, “Odensheite,” in Skrifter utgit av Videnskapsselskapet i Kristiania, 1924, II, Hist.-fil. Kl. No. 10. s. v., G. T. Flom, Journal of American Folklore XXXVIII (1925), 409. Has the OE name Brünstän a mythical significance?

^{Note 12 in page 943} Dagfinnr was a common and an ancient name which had been in use so long that this definition based on the meaning of its component themes is inept. A definition is given for the sake of completeness.

^{Note 13 in page 943} Hans Naumann, “Altnordische Namenstudien,” Acta Germanica, Neue Reihe, Heft I (Berlin, 1912), 85.

^{Note 14 in page 943} A. M. Sturtevant, Scandinavian Studies and Notes, IX (1927), 151 ff.

^{Note 15 in page 944} Sophus Bugge. Studier over de nordiske Gude— og Heltesagns Oprindelse (Christiania, 1881-89), 475, n. 3.

^{Note 16 in page 944} Detter, PBB XXI (1896), 107 n. 2.

^{Note 17 in page 944} This derivation is proposed by Finnur Jónsson in Afnf, XXXV (1909), 302. B. Sjöros in “Studier i nord. fil.,” III (1912) No. 2, (Skrifter utgivna av svenska litteratursällskapet i Finland CIII) connects Dúrnir with dyrr, ‘door,’ and gives it the meaning ‘doorkeeper.’ There is one story (Ynglingasaga, Chap. XII) where a dwarf is a doorkeeper, but the rather large number of names meaning ‘sleepy, lazy, drowsy’ indicates a strong tradition. One inclines to the meaning ‘sleeper’ though both are linguisticly possible.

^{Note 18 in page 944} Finnur Jónsson, Ordbog til de af samfund til udgivelse af gammel nordisk litteratur udgivne rimur samt til de af Dr. O. Jiriczek udgivne Bosarímur (København, 1926-27), s. v. dörr.

^{Note 19 in page 944} H. S. Falk and Alf Torp, Norwegisch-dänisches etymologisches Wörterbuch (Heidelberg, 1910-11), s. v. moden.

^{Note 20 in page 944} Alf Torp, Nynorsk etymologisk Ordbok, s. v. moden.

^{Note 21 in page 944} Elof Hellquist, Svensk etymologisk Ordbok (Lund, 1922), s. v. mogen.

^{Note 22 in page 945} Finnur Jónsson adopts the variant fíkmóinn in Skjaldedigtningen B I, 665.

^{Note 23 in page 945} Rudolf Meissner, Die Kenningar der Skalden (Bonn u. Leipzig, 1921), 154.

^{Note 24 in page 945} Hans Ross, Norsk Ordbog (Christiania, 1895) s. v. eikja; Bugge, FornkvœÐi, 93a.

^{Note 25 in page 945} Alexander Jóhannesson, Íslenzk tungu í fornöld (Reykjavik, 1924-24), §213, 3).

^{Note 26 in page 946} T. E. Karsten in “Studier i nord. fil.” IX (1928), No. 3, (Skrifter utgivna av svenska Litteratursällskapet i Finland CXXXIX). Otto v. Friesen, “Rö-stenen i Bohuslän,” Uppsala universitets årskrift, 1924. Filos. språkv. och hist. vetenskaper, 4, 86 f., 100.

^{Note 27 in page 946} Deutsche Altertumskunde (Berlin, 1883-91), V, 134.

^{Note 28 in page 946} Aisl. Gr. ⁴ §124, 2.

^{Note 29 in page 947} Gering here accuses Noreen of translating Fjalarr ‘(met)verstecker,’ in the third edition of the grammar, §119, 2, and explains at length that Fj$oolnir is the ‘metverstecker,’ which is exactly what Noreen says.

^{Note 30 in page 947} R. Keyser og P. A. Munch, Norges garnie Love indtil 1387 (Christiania, 1846-1895) I, 339.

^{Note 31 in page 947} idem. I, 363.

^{Note 32 in page 947} Afnf II (1885), 211 f.

^{Note 33 in page 948} Noreen Aisl. Gr. ⁴ §80, 2.

^{Note 34 in page 948} Elias Wessen, “Nordiska namnstudier,” Uppsala universitets Årsskrift, 1927, Filos, språkv. och hist. vetenskaper, 3, 90 n. 2.

^{Note 35 in page 948} The following words ending in Prim. Scand. -iR, some of them i-stems and some originally ju-stems, indicate noises; brestr, drynr, dykr, dynr, dynkr, fnyÐr, glymr, gnyÐr, gnyr, hlymr, hrinr, hrytr, hvinr, krytr, kviÐr, rymr, skellr, skr$oekr, stynr, styrr, svægr, ymr, ŷss, þrymr, þytr.

^{Note 36 in page 948} Sophus Bugge, Eelge-digtene i den œldre Edda, deres Hjem og Forbindelser (Kjøbenhavn, 1896), 326 n. 1.

^{Note 37 in page 948} To be found in FJ s. v. 2. Hr. For the various meanings and spellings often confounded with this word see Falk, Odensheite s. v.

^{Note 38 in page 949} Naumann, Altnordische Namenstudien, 35. Norenn, Aisl. Gr. ⁴, §291, 4.

^{Note 39 in page 949} As to the meaning of HerrauÐr and HerríÐr see note to Dagfinnr supra.

^{Note 40 in page 949} Fritzner, Ordbog over det gamle norske Sprog, s. v. hleÐi.

^{Note 41 in page 949} For a different derivation of Hlé- in proper names see Falk, Odensheite s. v Hléfréyr and Hléf$ooÐr.

^{Note 42 in page 949} Falk. Odensheite s. v.

^{Note 43 in page 950} Mangus Olsen, “Hedenske Kultminder i norske Stedsnavne,” Skrifter utgit af Videnskapsselskapet i Kristiania 1915, II, Histfil. Kl. No. 4, p. 237, n. 1.

^{Note 44 in page 950} FJ says Iaki is “vist kun forvansket af Iari.” Hermann Güntert, Kalypso (Halle 1919) 49, remarks on the derviation of jaki.

^{Note 45 in page 950} F. Detter u. R. Heinzel, Sœmundar Edda mit einem Anhang hrsgb. und erklärt (Leipzig, 1903) II, 22. Elof Hellquist, “Studier över de svenska sjönamnen” (Stockholm, 1903-6), 282, (Bidrag till kännedom om de svenska landsmålen och svenskt folkliv, XX, 1).

^{Note 46 in page 950} For comment on words similar to this in appearance, but not necessarily in origin or meaning, see Erik Brate, “Betydelsen av ortnamnet Skälv,” Namn och Bygd, I, (1913), 103 ff.

^{Note 47 in page 951} Gering (49) gives bibliography of derivations of Loki.

^{Note 48 in page 951} Sophus Bugge, FornkvœÐi 388. F. Detter, PBB, XXI (1886), 107 n. 1. Falk u. Torp, Norm.- dän. etymol. Wörterbuch (Heidelberg, 1910), s. v. suk.

^{Note 49 in page 951} Grimm, Deutsches Wörterbuch s. v.

^{Note 50 in page 951} Vigfusson, Icelandic-English Dictionary, s. v. m$oondull.

^{Note 51 in page 952} Occurs in þulur 2³. Both MSS have nœfi, which Finnur Jônsson prints nefi in Skjaldedigtningen B I, 672. This word is immediately preceded by Næfr in both MSS. Vide infra.

^{Note 52 in page 953} A. M. Sturtevant, “A Study of the Old Norse Word Reginn.” Journal of English and Germanic Philology, XV (1916), 251-266.

^{Note 53 in page 953} Finnur Jônsson, Afnf, XLIV (1928), 249.

^{Note 54 in page 954} Elias Wessen, Nordiske Namnstudier, 90, n. 3.

^{Note 55 in page 954} Sophus Bugge, Studier over de nordiske Gude- og Heltesagns Oprindelse (Christiania, 1881-89), 207 ff.

^{Note 56 in page 954} Eduard Sievers, PBB, XVIII (1899), 582 ff.

^{Note 57 in page 954} Two smiths, both named Varr, though not called dwarves, are to be understood as such. In Finnur Jônsson, Hrólfs saga kraka og Bjarkarímur (Kjøbenhavn. 1904), 15.

^{Note 58 in page 955} Nynorsk etym. Ordbok s. v.

^{Note 59 in page 955} E.g. þjóÐrerer, Finnur Jónsson, Eddalieder (Halle, 1888) 25. þjóÐrører, FJ s. v., Neckel, Edda ², 43. Gering (159). See Gering (129, 159) for references to older discussions.

^{Note 60 in page 956} Jôhannesson, op. cit. §272, 2.

^{Note 61 in page 956} Noreen, Aisl Gr. ⁴ 116. Codex regius of the Elder Edda has aan, Hauksbók has án in Vsp. 11⁷. One MS of the þulur (A. M. 748 II 4to) has the name Qnn. In the other MS this word is erased; possibly some one removed it who regarded it as an error.

^{Note 62 in page 956} They are the names of the children of a dwarf in a late novelized tale, “þorsteins saga Víkingssonar” in Rafn, FAS, II, 446-448. These two names are taken from the stock of old heroic names, and the parts are so fused that only a philologist could tell what each part means. When they are explained they tell nothing characteristic about dwarfs.

^{Note 63 in page 957} Names belonging to more than one group are repeated. E.g., compound adjective are listed among compounds among adjectives.

^{Note 64 in page 957} A. M. Sturtevant, “Old Norse tig -inn: tíg-inn; fú-inn: lú-inn,” Scand. S. and N. X (1928), 50-55

^{Note 65 in page 958} Finnur Jónsson, “Maskuline Substantiver på -nir,” Afnf, XXXV (1918), 302.

^{Note 66 in page 958} Frank Fischer, Lehnwörter des Altwestnordischen (Berlin, 1909), s. v. in index.

^{Note 67 in page 959} The intrusive endings -ill [two representatives] and -ull [one representative] are ignored here.

^{Note 68 in page 959} Fritz Wohlgemuth, Riesen und Zwerge in der afrz. erz. Dichtung (Tübingen, 1906), Diss.; August Lütjens, “Der Zwerg in der deutschen Heldendichtung des Mittelalters” Breslau, 1911), (Germanistische Abhandlungen hrsgb. von F. Vogt, Heft 38); Josef Riefel, Der Zwerg (Dresden-Weinbohl, 1923) contains much mateterial, no references; Helmut de Boor, “Der Zwerg in Skandinavien,” Festschrift, Eugen Mogk zum 70. Geburtstag (Halle, 1924), 536-581; C. N. Gould, “They who await the Second Death,” Scandinavian Studies and Notes, IX (1927), 167-201, and literature cited there, page 167, note 2.

^{Note 69 in page 959} C. W. von Sydow, Folkminnen och Folktankar, XII (1925), häfte 2, 15-16, doubts that the dwarves have anything to do with the dead. He treats the general subject of the relation of spirits and the dead in an article in the preceding häfte, pp. 1-10, and in XIII (1926), 172-174.

^{Note 70 in page 959} Hermann Güntert, Kalypso (Halle, 1919), 73 blár as color of corpse. K. E. Georges, Ausführliches lateinisch-deutsches Handwörterbuch (8th ed., Hannover and Leipzig, 1913) s. v. aquilus, ‘dark brown, blackish brown,’ ‘color of death and of the underworld.’

^{Note 71 in page 960} Certain names given the dwarves are also names of Othin, god of the dead; Brúni, Fj$oolsviÐr, Ginnarr, Grímr, Hár, Hrr, Reginn, þekkr, þrór. þrasir resembles the Othin-name þrasarr. Muninn, which is not a fitting dwarf-name, is an intruder from the Othin tales. Cf. Falk, Odensheite.

^{Note 72 in page 960} It is worth noting that a number of dwarf-names also appear as serpent-names in the Orma heiti, Skjaldedigtningen B I, 675. They are grímr, móinn, náinn, niÐhöggr, óinn, óri. The serpent-name haugvarÐr, ‘guardian of the mound,‘ is interesting in this connection. Güntert, Kalypso, 38. I. Reichborn-Kjennernd, “Ormen i nordisk Folkmedisin belyst ved den klassiske Oldtidsmedisin,” Tillægshæfte til Norges Apotekerforeningens Tidskrift, 1924, 16. It would be fruitless to include here the few dwarf-names which are also giant-names, horse-names and stag-names, since these categories have not been investigated.

^{Note 73 in page 960} Gould, op. cit. 172.

^{Note 74 in page 960} It is possible that the names indicating relationship and good qualities may be noa-terms for spiteful dwarves that would harm the person who spoke their real and evil names.

^{Note 75 in page 961} Lütjens op. cit. 103; Gould op. cit. 198 f.

^{Note 76 in page 961} Falk. Odensheite, s.v. pror.

^{Note 77 in page 961} E.g. Jón Helgason, HeiÐreks saga (København, 1914), 21. Carl Clemen, Religionsgeschichte Europas (Heidelberg, 1926), I, 231 remarks concerning Hephaistos “dasz eine aus der Erde aufsteigende Flamme auch sonst auf das mythische Feuer kleiner Erdgeister zurückgeführt wird.”

^{Note 78 in page 962} Gould, op. cit. 189.

^{Note 79 in page 962} Gould, op. cit. 199.

^{Note 80 in page 963} Lütjens, op. cit. 101.

^{Note 81 in page 963} The name Gandálfr is probably connected with magic or cult. Nils Lid, “Gand og Tyre,” Feslskrift til Hjalmar Falk (Oslo, 1927), 331 ff., shows that gand in MnN tradition may mean a magic ball of hair or other materials that can be put inside an animal to do it harm. Such a gand may be called an alvkule, ‘elf-ball,’ or dvergkule, ‘dwarf-ball,’ and may produce the illness known as alvskot or dvergskot.

A definite connection of the dwarf with magic is shown in Háv. 160, which tells of a dwarf chanting charms;

þat kann ek fimlânda, er gol þjóÐreyrir

dvergr, fyr Dellings durom:

afl gol hann ásom, en álfom frama,

hyggjo Hróptatý.

^{Note 82 in page 964} Elis Wadstein, “Norden och västeuropa i gammal tid” (Stockholm, 1925), (Populärt vetenskapliga föreläsninger vid Göteborgs Högskola, ny följd, XXII). See especially chapter X and the bibliography.

^{Note 83 in page 964} I quote them in the forms which Wadstein gives, op. cit. 148 f. kläde, skåt, duk, dok, kjortel, kyrtil, skräddare.

^{Note 84 in page 965} It is interesting to watch the transmitters of dwarf-names at the work of creating names. We are herewith chiefly concerned with eight manuscripts representing three streams of transmission, the first and second of which are closely related, though all three are from the same source. 1. R (gl. kgl. sml. n. 2365 4to (the Elder Edda) in the Royal Library at Copenhagen, and H (Hauksbók, cod. arnam. no. 544 4to in the Arnamagnaean collection in the University Library of Copenhagen) contain the V$ooluspá in versions that differ slightly from each other. 2. Manuscripts of the Younger Edda; T (Trektarbók, a paper manuscript in Utrecht); U (cod. upsaliensis no. 11, 8 vo. in the University Library in Upsala) W (codex Wormianus, cod. arnam. no. 242 fol in the University Library in Copenhagen); r (gl. kgl. sml. no. 2367 4to in the Royal Library in Copenhagen). These MSS contain the dwarf-name stanzas from the V$ooluspá in a somewhat different version from R and H. In a few places they agree with H as against R. 3. Manuscripts containing the Dverga heiti from the þulur; cod. arnam. 748 II 4to and cod. arnam 757 4to in the University Library at Copenhagen. The Dverga heiti consist of sixstanzas of eight short lines each, alliterative jingles composed of dwarf-names with almost no connective tissue.

One. could chart out the relationship of the manuscripts and discover which names were in the original list, the non-existent X back of all these streams.

But the disagreements are also interesting; they are of three sorts, one sort, rare, in which the scribe simply made an error and wrote something that does not exist in the Icelandic language, a second sort, also rare, where the transmitter failed to hear or to read the original correctly and put in a word that was an entirely good word but in no sense a dwarf-name, a third sort, rather frequent and very interesting, where some transmitter who failed to understand the original name made up and put in a good one in its place. All three sorts occurred with a name found in Vsp. 15⁴; TWr have HleÐiólfr, ‘door-wolf’ i.e. ‘shield-wolf,’ a good kenning for sword, and a suitable dwarf-name. The Dverga heiti have HljóÐólfr, ‘howl-wolf,’ also a good dwarf-name. But HleÐiólfr is probably the original because it is a kenning and harder to understand than HljóÐólfr, and the more difficult reading as a rule is the older. The innovation of the Dverga heiti was of the third sort. U has HleiÐólfr. HleiÐ- means nothing. It is simply a blunder, a change of the first sort. H has Hlévargr, ‘lee-wolf,’ i.e. ‘Iee-outlaw,’ which may well mean ‘the outlaw buried in the sheltered spot,’ a possible dwarf-name, but not a brilliant invention. This was again a change of the third sort. The transmitter remembered that the word had to alliterate with h and that it contained something about a wolf; he satisfied these conditions. It is very evident that oral transmission played a part here. R's Hlévangr must rest on faulty oral or written transmission of Hlévargr. Hlévangr, ‘lee-plain,’ i.e. ‘sheltered plain,’ is a normal word but in no sense a dwarf-name. It is a change of the second sort. The Dverga heiti contain a fairly good dwarf-name in Aurvargr, ‘gravel wolf,’ i.e. ‘outlaw buried in the gravel,’ but R and H have Aurvangr, ‘gravel-plain,’ in Vsp. 13⁷, a sensible word but one that tells nothing about dwarves. It is a change of the second sort from the original Aurvargr. In Vsp. 15³ R and H have Haugspori, UWr have Hugslari, 757 also has Hugstari and 748 has H$oogstari. All three are excellent dwarf-names, but Hugstari, being in MSS of two separate streams of transmission, was evidently the original name. These changes are of the third sort. In Vsp. 12¹ R has Veigr, ‘strength,’ decidedly suspect as a dwarf-name because it is the only abstract noun in the lot, H has Veggr, ‘wedge,’ U has Viggr, ‘axe-bit,’ Wr have vigr, ‘spear,’ three suitable dwarf-names. On account of the name Kíli, ‘wedge,’ one suspects that Veggr is here the original, but there is no proof. The change to Veigr was of the first sort, and if our suspicion is correct, those to Viggr and Vigr are of the third. There are other such changes, but these instances are sufficient to show that transmitters have again and again put in suitable new dwarf-names for old ones that were not well transmitted to them. The transmitters could draw independently from a fund of living tradition.

^{Note 85 in page 966} See n. 68 above.