The Icelandic alphabet (stafrof) in popular use as taught to children consists of the following letters (stafir):

a, b, c, d, e, f, g, h, i, k, l, m, n, o, p, q, r, s, t, u (v), x, y, z, þ, æ, ö,

the names of which may be learnt from two stanzas by Gunnar Pálsson in the Barna-gull:

A, bé, cé, dé, e, eff, gé,
eptir kemur há, í, ká,
ell, emm, enn, ó, einnig pé,
ætla eg par standi hjá.
Err, ess, té, ú eru þar næst,
ex, ý, zeta, þorn, æ, ö,—
allt stafrofið er svo læst
í erendin þessi lítil tvö.

The vowels are pronounced long. This alphabet was, with some additions, adopted from the Latin, and the þ was added at the end; and so late as the 17th century (in the Glossary of Magnus Ólafsson, who died 1636, and in the Icel. Grammar of Runolf Jónsson, who died 1654), the alphabet ends with þ, æ and ö being attached to a and o; Runolf calls the ö ‘o brevissimum.’ At a later time æ and ö were detached from a o, and put at the end; but not both of them at the same time, as Björn Halldorsson ends his Dictionary with æ. Gunnar Pálsson, who wrote the first popular abc, seems to be the man who, by his memorial stanzas, settled the alphabet as it is now taught. The division into mutes, liquids, etc. is too well known to be repeated. Neither are we here concerned with the Runic alphabet: there can be little doubt that this too was rudely imitated from the Greek or Latin, perhaps from coins: Roman coins of the 2nd and 3rd centuries of our era have been dug up in Scandinavian cairns and fens: foreign coined money was centuries in advance of books, and in barbarous countries shewed the way to the art of writing.

The vowels (hljóð-stafir or less properly raddar-stafir) are,

1. simple (short)—a, e, i, o, u, y, ö.

2. diphthongal, either marked with the acute (´), á, é, í, ó, ú, ý, or double letters, au, ei, ey, æ (œ).

Thus in written Icel. all the vowels together are, a á, e é, i í, o ó, u ú, y ý, æ, ö, the diphthongs au, ei, ey being included under a and e respectively. In this Dictionary the simple and acute vowels are treated under one head, but separately one after another; e.g. A in pp. 2-36, Á in pp. 36-48; these letters are widely different from one another both as to sound and etymology; a and á, o and ó, i and í, for instance, being no more akin than a and ei, o and au, etc.; and therefore great confusion would arise from mixing them together. The long vowels are chiefly due to contraction or absorption of consonants, which in Icel. has been carried farther than in any other Teutonic language, e.g. ar, atom, and ár, year; vin, friend, and vín, wine; dyr, door, and dýr, deer; fullr, full, and fúll, foul; goð, god, and góðr, good, etc.

To the consonants (samhljóðendr) were added in olden times the ð (eð), þ (þorn); and in modern times j, about the end of the last century; so that in Icel. writing all the consonants are, b, c, d, ð, f, g, h, j, k, l, m, n, p, q, r, s, t, v, x, z, þ, (=twenty-one); and this brings the whole alphabet to thirty-six letters:

a á, b, c, d ð, e é, f, g, h, i í, j, k, l, m, n, o ó, p, q, r, s, t, u ú, v, x, y ý, z, þ, æ (œ), ö,

from which number we may subtract c, q as little in use, x, z as compound letters, ð as subordinate to d, œ and æ are treated as one letter, and thirty remain; au, ei, ey go along with a and e, each in its due place, as also já, jó, jö, jú.

There is a curious division of the alphabet by an old Icel. grammarian of the latter part of the 12th century (Skálda 169-173). He draws five concentric circles: in the centre he places what he calls the höfuð-stafir (‘head-staves,’ initial letters), viz. h, q, v, þ, which in Icel. can only stand at the beginning of a syllable: in the next ring the mál-stafir (‘speech-staves’ or common consonants), twelve in number, which can stand both as final and initial: in the third ring the hljóð-stafir (‘voice-staves,’ vowels, still so called in Icel.), twelve in number, among which he distinguishes between six simple and six long vowels, the latter marked as at present with ´; with them also he counts the límingar (‘clusters,’ double vowels), æ, ꝏ, ꜹ, and lausa-klofar (split letters), ei, ey, as well as ia, io, iu; the vowel i he calls skiptingr (a changeling) from its being sometimes a vowel, sometimes a consonant: in the fourth ring are the capitals, which in MSS. are made to serve for double consonants (e.g. kroS = kross): lastly, in the fifth ring, the undir-stafir (‘under-staves,’ sub-letters), ð, x, z, which in Icel. can only be used as final.

Thorodd (Þóroddr Gamlason, called Rúnameistari or Rune-master) is the oldest Icel. grammarian, and lived in the beginning of the 12th century; for a curious account of this remarkable man, a builder by profession, see Bs. i. 235. He makes thirty-six vowels, nine of which seem to be nasal, caused by the frequent dropping and agglutination of n (in the infinitives, the weak nouns, etc.) These letters were lost before writing began, but left a nasal sound so late as the beginning of the 12th century. To the five Latin vowel characters he adds , , ø, y. These nine vowels as well as the nasals he then doubles by marking the long with an acute (´), and so they make thirty-six. In writing and printing, , , ø are out of use, but occur frequently in MSS.

Icel. prose literature extends over nearly eight centuries, and in the course of that time the language lost some of its rich vowel system; besides the nasals we are able to trace seven distinct vowels as lost. Four of them were lost at a very early time, perhaps in the 12th century, viz. ́ the umlaut of á (sec p. 1, B. 5); ø or œ, a vowel change of ó; and the double e and ö sound (see introduction to letter E); all these four letters were lost about the same time, and so early that few MSS. use them; they are not noticed in this Dictionary, except now and then for etymological purposes. Some three or four centuries later, three other vowels vanished, viz. the y sound in all the three letters y, ý, ey, which became respectively = i, í, ei; but the former are still preserved in writing and printing. The MSS. down to the Reformation make in most cases a sharp distinction between the i and y sound, as also the poets; yet one very ancient MS. of the 12th century (Arna-Magn. 623, see Frump, pp. 42-48) is remarkable for its confounding both letters. The same confusion is observable in Anglo-Saxon; whereas in Denmark, Sweden, and Norway, the distinction of i and y is still strictly kept up. As for Icel. we suspect that the change began in some remote district at an early time, until many centuries later it was suddenly adopted throughout the whole country.

The Icel. is not, in its pronunciation, a strongly accented language, (the acutes, as stated above, are marks of diphthongs, not of accent,) and is in this respect nearest in sound to the French. In modulation the Icel. is in the main trochaic (– ⏑ | – ⏑), and arsis and thesis follow alternately one after another: secondly, all root syllables are accentuated, but inflexive syllables have no accent, e.g. bārnă, hāndă, bōðă, hārðăn, fāgră; in bisyllabic compounds both the root syllables are accentuated, but the second with only a half accent, which we mark by ⏓, e.g. sām-bā̆nd, hūg-bō̆ð, as also in strong inflexions like -andi, -astr, e.g. eīgā̆ndĭ, hārðā̆stăn: if one of the words which form a compound falls in the third syllable it is accentuated, e.g. bārnă-gūll, bārnă-gūllĭ, hāndă-vērk (but hānd-vē̆rk), because in this case the arsis falls on the third syllable which is a root: in trisyllabic words with bisyllabic inflexion the third syllable is sounded ⏓, e.g. laūsnărī̆n, hūggŭnā̆r, sȳngăðī̆st, sāmleĭkī̆nn, hēntŭgā̆st, trūăr-ī̆nnăr (fidei), nāðăr-ī̆nnăr, hōfðĭng-jā̆nnă, and that even though the second syllable is a root syllable, e.g. ūppvā̆knā̆ðr, āfsō̆kū̆n: words like blēssŭnăr|ī̆nnăr, mīskŭnăr|ī̆nnĭr, drōttnĭngăr|ī̆nnăr, etc. are dactylic. Root and inflexion on the one hand and the trochaic flow on the other are felt all along, mutually resisting or aiding one another as to the measure of a syllable; accordingly, whenever the arsis falls on ⏑ it becomes ⏓, if on ⏓ it becomes –. In the best Icel. poets half-accentuated syllables may form full rhyme, by a poetical licence; thus, in the Passíu-Sálmar more than eight score, and in Búnaðar-balkr more than two score of such rhymes are found, e.g.

Mig hefir ljúfur Lausnarinn
leitt inn í náðar grasgarð sinn.
Huggun er manni mönnum að
miskun Guðs hefir svo tilskikkað.
Iðranin blíðkar aptur Guð
ei verður syndin tilreiknuð.

Bænarlaus aldrei byrjuð sé
burtför af þínu heimile.
Þu veizt ei hvern þú hittir þar
heldur en þessir Gyðingar.
Því hjartað mitt er helmingað,
hlakka eg til að finna það.


A is the first letter in all the alphabets of Phenician extraction. The Runic alphabet, being confused and arbitrary, makes the sole exception to this rule.

A. Pronunciation: it is either simple (a) or diphthongal (á). The simple a is pronounced long or short; when long it is sounded like the long Italian a as in padre, or as in Engl. father; when short, like the short Italian a as in cambio, or as in Engl. marry. The á—though in grammars commonly called a long vowel—is phonetically diphthongal (a + u), and sounds like Engl. ou or ow: Engl. thou and Icel. þá, now and ná, have almost the same sound. Again a and á have, like all other vowels, diphthongs or simple, a deep, full chest-sound if followed by a single consonant, or by more than one weak consonant (a liquid followed by a media). They sound short if followed by two or more strong consonants (a double mute or liquid): thus the a and á sound long in tāl, sermo; sāt, sedebat; mān, mancipium; tā́l, dolus; ā́r, remus; sā́t, sessio, hātr, odium; hārðr, durus; kāldr, frigidus; vāndr, difficilis; tāmdr, domitus, etc. But short in hătt, pileum; hắtt, modum; mănn, hominem; bănn, interdictum; hắll, lubricus; kălt, frigidum; rămt, acidum; hărt, durum; vănt, assuetum, etc.; the consonants shortening the sound of the preceding vowel. The a is also short in all endings, verbal or nominal, tālă, tālăr, tālăðă, dixi; tālăst, dicitur; vākă, vigilia; fāgrăn, pulchrum, etc. Etymologically a distinction must be made between the primitive á, as in sátu (sedebant), átu (edebant), gátu (poterant), and the á produced by suppressing consonants; either nasals, as in á, ást, áss, báss, gás, = an, anst, ans, bans, gans; or gutturals, h, g, k, as in á (aqua), (videbat), (jacebat), (debet), nátt (nox), dráttr (tractus), and a great many others; or labials, v, f, as in á = af, áir = afr, hár but háfan; or dentals, as in nál (acus) [Goth. neþla, Engl. needle], vál (ambitus, mendicitas) [A.S. vädl], etc. In very early times there was no doubt an audible distinction between these two kinds of á, which however is not observed even by the earliest poets, those of the 10th century. The marking of the diphthongal vowels with an acute accent is due to the Icelandic philologist Thorodd (circa 1080–1140), and was probably an imitation of Anglo-Saxon. The circumflex, applied by Jacob Grimm, is unknown to Icel. authors of whatever age. Thorodd, in his treatise on the vowels (Skálda, pp. 160 sqq.), distinguishes between three kinds of vowels, viz. short, long (i.e. diphthongal), and nasal. The long ones he proposes to mark with an acute (´); the nasals by a dot above the line (˙). The vowels of his alphabet are thirty-six in number. According to his rule we should have to write, af (ex), át (esus), ȧ (in). No doubt the a was also nasal in the verbs and the weak nouns, komȧ (= koman), augȧ (gen.); and also when followed by an n, e.g. vȧnr (assuefactus). The distinctive marking of the nasals never came into practice, and their proper sound also disappeared; neither is this distinction observed by the poets in their rhymes. The marking of the diphthongal vowels—either the primitive vowels or those formed by agglutination—by an acute accent, according to the rule of Thorodd, is indeed used in a very few old Icel. parchment fragments of the 12th century. The only MS. of any considerable length which strictly observes this distinction is the Ann. Reg. Ísl. 2087. 4b. Royal Libr. Copenhagen, written in Icel. at the end of the 13th century. In the great bulk of MSS. both kinds of vowels are treated alike, as in Latin. About the middle of the 14th century the doubling of vowels, especially that of aa () = á, came into use, and was employed through more than three centuries, until about 1770 the Icelanders resumed the spelling of Thorodd, marking diphthongal vowels by an acute accent, but following the rules of modern pronunciation. The diphthong au—in Norse freq. spelt ou—has at present in Icel. a peculiar sound, answering to äu or eu in German, and nearly to Engl. oi. The Norse pronunciation is different and perhaps more genuine.

B. Changes.

I. a changes into e, á into æ: this change—a part of a more general transformation, by Grimm termed umlaut, ‘vowel-change’—is common to all the Teutonic idioms, except the Gothic (v. letter E and Æ).

II. a changes into ö (), á into ́: this transformation is peculiar to the Scandinavian branch, esp. the Icelandic idiom, where it is carried on to the fullest extent—in old Swedish and Danish its use was scanty and limited. It takes place,

1. in monosyllabic nouns with a for their radical vowel,

α. feminines, öld, periodus; önd, anima; örk, arca; för, iter; höll, aula; hönd, manus; sök, causa, etc.

β. adjectives in fem. sing. and in neut. pl., öll, tota; fögr, pulchra; hörð, dura; hölt, clauda; sönn, vera; from allr, etc.

γ. in plur. neut., bönd, vincula; börn, τέκνα; lönd, terrae; from band, etc.

δ. in singular masculines with a suppressed u in the root, hjörtr, cervus; fjörðr, sinus; björn, ursus; örn, aquila, etc.

2. in dissyllables a radical a, when followed by a final u (-u, -ur, -um, etc.), in Icel. constantly changes into ö,öllum, cunctis; mönnum, hominibus; köllum, vocamus; vökum, vigiliis and vigilamus; vökur, vigiliae, etc. Danes and Swedes here retained the a; so did a great part of Norway. The change only prevailed in the west of Norway and the whole of Iceland. Some Norse MSS. therefore constantly keep a in those cases, e.g. Cd. Ups. De la Gard. 8 (Ed. C.R. Unger, 1849), which spells allum, cunctis; hafuð, caput; jafur, rex; andverðr, adversus; afund, invidia, etc. (v. Pref. viii.) Other Norse MSS. spell a and ö promiscuously; allum or öllum, kallum or köllum. In Icel. this change prevailed about the year 1000. Even at the end of the 10th century we still frequently meet with rhymes such as barð—jarðu, þang—langu, etc.

3. a in inflexions, in penultimate syllables, if followed by u, changes into u (or ö); thus keisurum, caesaribus; vitrurum, sapientioribus; hörðurum, durioribus; hörðustum, durissimis: pret. pl., sköpuðu, creabant; töluðu, dicebant; orrustu, pugnam. In part. pass. fem. sing. and neut. pl., sköpuð, creata; töluð, dicta; töpuð, perdita. Neut. pl. in words, as sumur, aestates; heruð, pagi. This change is peculiar to Iceland, and is altogether strange to Norse MSS., where we constantly find such forms as ætlaðu, putabant; gnagaðu, mordebant; aukaðu, augebant; skapað, creata; kallað, dicta; skaparum, tapaðum, ágætastum, harðarum, skínandum; kunnastu, artem, etc. This difference, as it frequently occurred at early times, soon gave the Icel. idiom a peculiar and strange sound,—amarunt would, in Icelandic, be ömurunt. Norse phrases—as með bænum ok fastu (föstu) hafðu (höfðu) með sér vaxljós, ok dýrkaðu (dýrkuðu) þa hælgu hátíð með fastu (föstu) ok vaktu (vöktu) þar um nóttina með margum (mörgum) aðrum (öðrum) vanfærum mannum (mönnum), O.H.L. 87—sound uncouth and strange to Icel. ears; and so no doubt did the Icel. vowel transformations to Norse ears.

4. endings in -an, -all, e.g. feminines in -an, as hugsan, ætlan, iðran, frequently change into -un,—hugsun, ætlun, iðrun, and are now always used so: gamall, vetus, f. gömul; einsamall, solus, f. einsömul. In modern Norse, gomol, eismol (Ivar Aasen); atall, atrox; ötull, strenuus; svikall, perfidus, and svikull; þrifnaðr, mundities, and þrifnuðr, etc.

5. in the cases correlative to II. 1, 2, the á in its turn changes into a vowel, by Thorodd marked ́; this vowel change seems to have been settled about the beginning of the 11th century, and prevailed in Iceland during the 12th,* this vowel change seems still to have been in full use in Icel. during the 11th and 12th centuries, being constantly employed in MSS. of that time; about the end of that century, however, and the beginning of the next, it fell off, and at last became extinct. Its phonetical value, therefore, cannot now be precisely stated: it no doubt had an intermediate sound between á and ó, such as ö () has between a and o. Thorodd proposed to mark the short ‘umlaut’ ö by ; and the vowel change of á by ́ (in the MSS. however commonly written ǫ).

Instances: fem., ́, amnis; ́st, amor; ́l, funis; ́r, remus; l́g, lignum; skr́, libellus; s́tt, pax; s́l, anima; n́l, acus; v́n, spes: masc., h́ttr, modus; þr́ðr, filum; þ́ttr, funis; m́ttr, vis; ́ss, deus; ́rr, nuntius: neut. pl., s́r, vulnera; t́r, δάκρυα; m́l, dicta; r́ð, consilia; v́r, vera: adj. fem, and neut., k́t, læta; f́, pauca; sm́, parva; h́, alta; f́m, paucis; h́m, altis: verbs, s́, videbant (but , videbat); g́tu, capiebant; ́tu, edebant (but át, edebat), etc.: v. Frump. 26–28: e.g. sár (vulnus) veitti maðr mér eitt (unum), s́r mörg (multa vulnera) veitta ek hánum, Skálda (Thorodd), 162; l (= öl, cerevisia) er drykkr, ́l er band (vinculum), id. 163; tungan er málinu vn (= vön, assuefacta), en at tönnunum er bitsins v́n (morsus exspectatio), id.: frequently in the Grágás, lýsa sár sitt (vulnus) eðr s́r (vulnera) ef fleiri eru, Kb. i. 151; s́r en minni (vulnera leviora), 170; en meire s́r (graviora), 174; síðan es s́r eða ben voru lýst, 175; engi s́r (nulla vulnera), sr, and rð, 176, 177; m́l, ii. 51; v́r, 158, etc.

C. Other Changes:—in modern Icel. the old syllable has changed into vo; vó of the 14th century being an intermediate form: thus von, spes; votr, madidus; vor, ver; vorr, noster; voði, periculum; koma, adventus; voru, erant, etc.: so also the á in the dat. hánum, illi, now honum, which is also employed in the editions of old writings; kómu = kvámu = kvómu, veniebant, etc. In Norway a was often changed into æ in the pronominal and adverbial forms; as hæna, illam; þær, þænn, þæt, ibi, ilium, illud; hence originate the mod. Dan. hende, der, den, det; in some Norse dialects even still dar, dat. The short a in endings in mod. Dan. changed into e (æ), e.g. komme, uge, talede, Icel. koma, vika; whereas the Swedes still preserve the simple a, which makes their language more euphonious than the mod. Dan. In most districts of Icel. an a before ng, nk, has changed into á, thus langr (longus), strangr (durus), krankr (aegrotus) are spelt lángr, kránkr, etc. In the west of Iceland however we still say langr, strangr, etc., which is the pure old form. The a becomes long when followed by lf, lm, lp, thus álfr, genius; álpt, cygnus; hálfr, dimidius; kálfr, vitulus; sjálfr, ipse; this is very old: the fem. hlf, dimidia, which occurs in the 12th century, points to an á, not a; já = ja in hjálpa, skjálfa, etc. The lengthening before lm is later,—álmr, ulmus; hálmr, calamus; sálmr, psalmus; hjálmr, galea; málmr, metallum, etc. In all these cases the á is not etymological. Also before ln in the plur. of alin, álnar not alnar: lk, álka = alka, alca; bálkr = balkr; fálki = falki, falco: háls = hals; frjáls = frjals; járn = jarn; skáld = skald; v. those words: aarni, dat. of arinn, v. that word: the proper name Árni, properly Arni: abbati, abbas, ábóti: Adám, on the contrary, changed into Adam; Máría into Maria, Mary. The old spelling is still kept in máriatla, motacilla pectore albo, etc. In the 1st pers. pret. indic., and in the pres. and pret. conj. we have a changed into i, e.g. talaða to talaði, locutus sum; sagða, dixi, vilda, volui, hafða, habui, to sagði, vildi, hafði: in the 1st pers. pres. and pret. conj., hefða, haberem, hafa, habeam, to hefði, hafi. These forms occur as early as the beginning of the 13th century (e.g. in the Hulda, Cd. A.M. 66, fol. = Fms. vi. and vii). In the south of Iceland however (Reykjavík, the Árnes and Gullbringusýsla) the old forms are still frequently heard in bisyllabic preterites, esp. ek vilda, sagða, hafða, and are also employed in writing by natives of those districts.

D. a answers to Goth. a; A.S. ea (a, ä); allr, totus; Goth. alls; A.S. eall: the primitive á to Goth. ê, sátu, Goth. sêtun, sedebant; gráta, grêtun, lacrymari; láta, lêtan; vápn, vêpn, arma; vágr, vêgs, fluctus. The Icel. secondary á, on the contrary, must in the kindred Teutonic idioms be sought for under a vowel plus a consonant, such as an, ah, or the like. A.S. æ commonly answers to Icel. á, láta, A.S. lætan; dáð, A.S. dæð; þráðr, A.S. þræð, Engl. thread; mál (καιρός), A.S. mæl, cp. Engl. meal. The A.S. â, on the contrary, etymologically answers to Icel. ei. The diphthong au answers to Goth. au, A.S. ea,—rauðr, Goth. rauds, A.S. reað, Engl. red. In English the a seems at very early times to have assumed its present ambiguous sound; this we may infer from A.S. words introduced into Icelandic. The river Thames in Icel. is spelt, as it is still pronounced in England, as Tems, which form occurs in a poem of the year 1016.

E. The Runic character for a was in the Gothic and Anglo-Saxon Runes (so termed by P.A. Munch) [A.S. ]; so in the Golden horn, on the stone in Thune in Norway (Ed. by P.A. Munch, 1857), and in the Bracteats. The Saxons called it ôs = áss, deus. In the Runes it was the fourth letter in the first group (fuþork). The Scandinavians in their Runes used this character for o, and called it óss, ostium, probably misled by the A.S. pronunciation of the homely word áss. This character, however, occurs only a few times in the common Runes, which in its stead used the A.S. Rune j, gêr, annona, which is the fourth Rune in the second group (hnias, A.S. hnijs), called according to the northern pronunciation ár, annona: this letter, or has the form, as well as the name and place, of the A.S. j, .


B (bé) is the second letter. In the Phenician (Hebrew) alphabet the three middle mutes, b, g, d, etc., follow in unbroken order after a. In the Greek the same order is kept; in Latin, and hence in all European alphabets, a confusion arose, first, by giving to the gamma (the old Greek gamma) the value of k (c), and thereby throwing g out of its original place: secondly, by placing e and F (identical in form with Ϝ, the old Greek digamma) immediately after the d; thus, instead of the old Greek (and Hebrew) a, b, g, d, e, f, we got a, b, c, d, e, f, g, etc. In the old Slavonian alphabet v (vidil) was inserted between the b and g (Grimm Introd. to lit. B). In the old Runic alphabet the order became still more disjointed; the common rude Scandinavian Runes have no special g or d, and their b is put between t and l, nearly at the end of the alphabet (… t, b, l, m, y). In all the others b kept its place at the head of the consonants, immediately after a, which stands first in almost all alphabets.

A. Among the vowels a begins more words than any other vowel: it contains the three great prepositions, af, at, and á, which, with their compounds, along with those of al- and all-, make up more than half the extent of the letter; it abounds in compound words, but is comparatively poor in primitive root words. Again, b is in extent only surpassed by the consonants h and s; in regard to the number of root words it is equal to them all, if not the foremost. It is scanty in compounds, has no prepositions, but contains the roots of several large families of words, as, for instance, the three great verbs, bera, bregða, and búa; besides many of secondary extent, as binda, bíða, biðja, etc.; and a great number of nouns. The extent of b is greatly reduced by the fact, that the Scandinavian idioms have no prefix be-, which in the German swells the vocabulary by thousands (in Grimm it takes up about 300 pages); the modern Swedes and Danes have during the last few centuries introduced a great many of these from modern German; the Icel. have up to the present time kept their tongue pure from this innovation, except in two or three words, such as betala or bítala (to pay), befala or bífala (to commend), behalda or bíhalda (to keep), which may, since the Reformation, be found in theol. writers; the absence of the prefix be- is indeed one of the chief characteristics of the Icel. as opposed to the German; the English, influenced by the northern idiom, has to a great extent cut off this be-, which abounds in A.S. (v. Bosworth, A.S. Dictionary, where about 600 such words are recorded); even in the Ormulum only about thirty such words are found; in South-English they are more frequent, but are gradually disappearing. Again, b represents p in Scandinavian roots; for probably all words and syllables beginning with p are of foreign extraction; and the same is probably the case in German and English, and all the branches of the Teutonic (vide Grimm D.G. iii. 414); whereas, in Greek and Latin, p is the chief letter, containing about a seventh of the vocabulary, while b contains from one seventieth to one ninetieth only. It might even be suggested that the words beginning with b in Greek and Latin are (as those with p in the Teutonic) either aliens, onomatopoëtics, provincialisms, or even cant words.

B. Pronunciation.—The b is in Icel. sounded exactly as in English:

I. as initial it is an agreeable sound in all the branches of the Teutonic, especially in the combinations br and bl, as in ‘bread, brother, bride, bloom, blithe, blood, bless,’ etc. etc. The Greek and Roman, on the other hand, disliked the initial b sound; but the difference seems to be addressed to the eye rather than the ear, as the π in modern Greek is sounded exactly as Icel. b, whilst β is sounded as Icel. v; thus the Greek βίσων in Icel. rendered phonetically by vísundr, but ἐπίσκοπος (biskup, bishop) is in all Teutonic dialects rendered by b, not p, probably because the Greek π had exactly this sound.

II. but although agreeable as the initial to a syllable, yet as a middle or final letter b in Icel. sounds uncouth and common, and is sparingly used:

1. after a vowel, or between two vowels, b is never sounded in Icel. as in modern German geben, haben, laub, leben, leib, lieb; in all those cases the Icel. spells with an f, sounded as a v. Ulfilas frequently uses b, e.g. graban, haban, saban, ïba, gabei, etc.; yet in many cases he vacillates, e.g. giban, graban, gêban, grôbun, tvalib, but gaf and grôf, etc. So gahalaiban on the Gothic-Runic stone in Tune, but hlaifs, Ulf., Luke vi. 48. The Greek and Latin abound in the use of the b (bh) in the middle of syllables and inflexions (-bus, -bills, -bo): in Icel. only a double b may be tolerated, but only in onomatopoëtic or uncouth words, as babbi (pa of a baby), bobbi (a scrape), stubbi (Germ. stumpf), lubbi (Germ. lump), nabbi (a knob), krabbi (a crab), gabb, babbl, babbla, etc.; cp. similar words in English.

2. joined to a consonant;

α. in old Swedish b is inserted between m and r or m and l (as in mod. Greek μρ and μλ are sounded μβρ and μβλ, e.g. Swed. domber, komber, warmber, hambri, gamblar = Icel. dómr, kömr (venit), varmr, hamri, gamlar: Swed. kumbl and kubl (Icel. kuml, monumentum) are used indifferently. Even in old Icel. poems we find sumbl = suml, symposium, simbli = simli, Edda i. 256 (Ed. Havn.): mp is only found in adopted words, as in kempa (cp. Germ. kampf), lampi (Lat. lampas), and is almost assimilated into pp (kappi): mb is tolerated in a few words, such as umb, lamb, dramb, dumbr, kambr, vömb, timbr, gymbr, strambr, klömbr; cp. the Engl. lamb, comb, timber, womb, where the b is not pronounced (except in the word timber); in limb, numb the b is not organic (cp. Icel. limr, numinn); it occurs also in a few diminutive pet names of children, Simbi = Sigmundr, Imba = Ingibjörg. In the 16th and 17th centuries the Germans used much to write mb or mp before d or t, as sambt or sampt (una cum), kombt or kompt (venit); but this spelling again became obsolete.

β. the modern High German spells and pronounces rb and lb, werben, korb, kalb, halb, etc., where the middle High German has rw and lw, korw, kalw; the modern Scandinavian idioms here spell and pronounce rf, lf, or rv, lv, e.g. Dan. kalv, Swed. kalf, vitulus; the Icel. spells with f, arfi, kálfr, but pronounces f like v. Yet in Icel. rb, lb are found in a few old MSS., especially the chief MS. (A.M. folio 107) of the Landnáma, and now and then in the Sturlunga and Edda: nay, even to our own time a few people from western Icel. speak so, and some authors of mark use it in their writings, such as the lexicographer Björn Halldórsson, e.g. álbr, kálbr, hálbr, sjálbr, silbr, úlbr, kólbr, orb, arbi, karbi, þörb, = álfr, etc.; only the word úlbúð, qs. úlfúð, is used all over Icel.

γ. fl and fn are in mod. Icel. usage pronounced bl and bn, skafl, tafl, nafli, = skabl, tabl, nabli; nafn, höfn, safn, nefna, = nabn, höbn, sabn, nebna; without regard whether the radical consonant be f or m, as in nafn and safn, qs. namn and samn. This pronunciation is in Icel. purely modern, no traces thereof are found in old vellum MSS.; the modern Swedes, Danes, and Norse pronounce either mn (the Swedes spell mn where Icel. use fn or bn) or vl (Dan.), ffl (Swed.)

δ. is in Icel. commonly pronounced as bð, e.g. hafði, hefð, sofðu = habði, hebð, sobðu; yet a few people in the west still preserve the old and genuine pronunciation vd (havdu, sovdu, not habðu, sobðu), even in the phrase, ef þú (si tu), proncd. ebðú. The prefixed particles of- and af- are often in common speech sounded as ob-, ab-, if prefixed to a word beginning with b or even m, l, e.g. ofboð, afburðr, afbindi, aflagi, afmán, as obboð, abbindi, Hm. 138; abbúð, Korm. 116; abburðr, Fms. x. 321; ablag, abmán: gef mér, lofa mér, proncd. gébmér or gémmér, lobmér or lommér (da mihi, permitte mihi); af mér (a me), proncd. abmér or ammér; but only in common language, and never spelt so; cp. Sunnan Póstur, a.d. 1836, p. 180, note **.

ε. b = m in marbendill = marmennill.

C.According to Grimm’s Law of Interchange (‘Lautverschiebung’), if we place the mute consonants in a triangle thus:
f→b→p→f b→g→k→b þ→d→t→þ
the Scandinavian and Saxon-Teutonic form of a Greek-Latin root word is to be sought for under the next letter following the course of the sun; thus the Greek-Latin f (φ) answers to Icel. and Teutonic b; the Greek-Latin b (β), on the other hand, to Teutonic p. Few letters present so many connections, as our b (initial) does to the Greek-Latin f, either in whole families or single words; some of the instances are dubious, many clear: φάλαγξ, cp. Icel. balkr; φάρ, Lat. far, cp. barr; φαρόω, φάρος, Lat. fŏrare, cp. bora; φάρυγξ, cp. barki; φόβος, φοβέω, cp. bifa; φέρω, φορέω, Lat. fĕro, cp. bera, borinn; φόρτος, cp. byrðr; φεύγω, εφυγον, Lat. fŭgio, cp. beygja, boginn, bugr; φηγός, Lat. fāgus, cp. bók, beyki; φλέγω, φλόξ, Lat. fulgere, fulgur, cp. blik, blika; φλέω, Lat. flāre, cp. blása, bólginn, Lat. follis, cp. belgr; φλογμός, Lat. flōs, cp. blóm; φονή, φόνος, φεν-, cp. bani, ben; φορμός, cp. barmr; φράγμα, φράσσω, cp. borg, byrgja; φράζω, φραδή, cp. birta; φρατήρ, Lat. frāter, cp. bróðir; φρέαρ, cp. brunnr; φρίσσω, cp. brattr (brant), brandr; ὄφρυς, cp. brá; φρύγανον, φρύγω, cp. brúk; φύω, Lat. fīo, fŭi, cp. búa, bjó, Engl. to be, and the particle be- (v. Grimm s.v. be- and bauen); φύλλον, Lat. fŏlium, cp. blað; φώγω, Lat. fŏcus, cp. baka: moreover the Lat. făcio, -fĭcio, cp. byggja; fastigium, cp. bust; favilla, cp. bál; fĕrio, cp. berja; fĕrox, fĕrus, cp. ber-, björn; fervere, cp. brenna; fīdus, foedus, cp. binda; findo, fīdi, cp. bíta, beit; flăgellum, cp. blaka; flectere, cp. bregða; fluctus, cp. bylgja; fŏdio, cp. bauta, Engl. to beat; fundus, cp. botn; fors, forte, cp. ‘burðr’ in ‘at burðr;’ frango, frēgi, frăgor, cp. breki, brak, brjóta; fraus (fraudis), cp. brjóta, braut; frūges, fructus, cp. björk; fulcio, cp. búlki; frĕmo, cp. brim; frenum, cp. beisl, Engl. bridle; frons (frondis), cp. brum;—even frons (frontis) might be compared to Icel. brandr and brattr, cp. such phrases as frontati lapides;fātum, fāma, cp. boð, boða, etc. The Greek φίλος, φιλεîν might also be identical to our bl- in blíðr. The change is irregular in words such as Lat. pangere, Icel. banga; petere = biðja; parcere = bjarga; porcus = börgr; πηγή, cp. bekkr; probably owing to some link being lost.

β. in words imported either from Greek or Roman idioms the f sometimes remains unchanged; as the Byz. Greek φεγγάριον is fengari, Edda (Gl.); sometimes the common rule is reversed, and the Latin or Greek p becomes b, as episcopus = biskup; leopardus = hlébarðr, Old Engl. libbard; ampulla = bolli; cp. also Germ. platz = Icel. blettr; again, plank is in the west of Icel. sounded blanki: on the other hand, Latin words such as bracca, burgus are probably of Teutonic or Celtic origin.

γ. the old High German carried this interchange of consonants still farther; but in modern High German this interchange remains only in the series of dental mutes: in the b and g series of mutes only a few words remain, as Germ. pracht (qs. bracht), cp. Engl. bright; Germ. pfand, cp. Engl. bond; otherwise the modern Germans (High and Low) have, just as the English have, their braut, bruder, brod, and butter, not as in old times, prût, etc.

D. In the Runic inscriptions the b is either formed as B rune, so in the old Gothic stone in Tune, or more commonly and more rudely as in the Scandinavian monuments; both forms clearly originate from the Greek-Roman. The Runic name was in A.S. beorc, i.e. a birch, Lat. betula; ‘beorc byð blêda leâs …,’ the A.S. Runic Poem. The Scandinavian name is, curiously enough—instead of björk, f. a birch, as we should expect—bjarkan, n.; the name is in the old Norse Runic Poem denoted by the phrase, bjarkan er lauf grænst lima, the b. has the greenest leaves, cp. also Skálda 177: both form and gender are strange and uncouth, and point to some foreign source; we do not know the Gothic name for it, neither is the Gothic word for the birch (betula) on record, but analogously to airþa, hairda, Icel. jörð, hjörð, björk would in Gothic be sounded bairca, f.; the Scandinavian form of the name points evidently to the Gothic, as a corruption from that language,—a fresh evidence to the hypothesis of the late historian P.A. Munch, and in concord with the notion of Jornandes, about the abode of the Goths in Scandinavia at early times. Thorodd (Skálda 166) intended to use b as a sign for the single letter, B for a double b, and thus wrote uBi = ubbi; but this spelling was never agreed to.


C (cé), the third letter, has all along been waning in Icel. The early Gothic Runes (Golden horn) use for k, e.g. ᛖᚲ for ek, ego; the later common Runes have no c. The Anglo-Saxon Runes follow the Gothic, and use c for k, as cén, a torch.

A. Spelling.—The rule given by the first Icel. grammarian, Thorodd (a.d. 1140), is curious; he says that he will follow the Scots in using c with all the vowels, as in Latin, and then makes c serve instead of k; but, though in other cases he makes the small capitals serve for double consonants, e.g. uBi, braT, meN, haLar, döG, = ubbi, bratt, menn, etc., he admits k to mark a double c, and spells söc sake, but sök sank; lycia to shut, but lykja a knot; vaca to wake, but vaka vagari; þecia to thatch, but þekia to know. Thorodd gives as his reason that other consonants have different shapes as small or capital, but c is uniform, whereas he says that k suits well for a double c, being a Greek letter itself, and having a shape similar to a double c, namely, K as |(; this k or double c he calls ecc, but the single c he calls ce, Skálda 108. The second grammarian (about the end of the 12th century) only admits c as a final letter, ranking with ð, z, or x, which are never used as initials: all these letters he calls ‘sub-letters;’ he thus writes karl, kona, kunna, but vöc, söc, tac. Such were the grammatical rules, but in practice they were never strictly followed. As the Anglo-Saxon, in imitation of the Latin, used c throughout for k, so the earliest Icel. MSS., influenced by the Anglo-Saxon or by MSS. written in Britain, made free use of it, and k and c appear indiscriminately; k is more frequent, but c is often used between two vowels or after a vowel, e.g. taca, lécu, vica, hoc, etc. etc. In such cases, t and c (k) can often hardly be distinguished; and readings can sometimes be restored by bearing this in mind, e.g. in Bjarn. S. (all our MSS. come from a single vellum MS.) the passage ‘létu heim at landinu’ should be read ‘lécu (léku) honum landmunir,’ 16; ‘sáttvarr’ is ‘sacvarr,’ i.e. sakvarr, 51; cp. also such readings as bikdælir instead of Hitdælir, Gullþ. 3; drickin = dritkinn, id. In Ad. 20 it is uncertain whether we are to read veclinga- or vetlinga-tös, probably the former.

B. Foreign words.—Throughout the Middle Ages the spelling remained unsettled, but k gained ground, and at the time of the Reformation, when printing began, c was only kept to mark the double k, ek (cut on one face), and in foreign proper names; but it was not admitted in appellatives such as kirkja, klaustr, klerkr, kór, kross, kalkr or kaleikr, church (Scot. kirk), cloister, clericus, choir, cross, calix, etc., or in kista, kastali, kerti, keisari, kær, kærleiki, kyndill, kórona or krúna, kurteisi, kumpan, kompás, kapítuli, cista, castellum, cern, caesar (as appell.), carus, caritas, candela, corona, courtesy, company, compass, chapter. All words of that kind are spelt as if they were indigenous. The name of Christ is usually in editions of the N.T. and Vídal. spelt Christus or Christur, but is always sounded as a native word Kristr or Kristur, gen. Krists, dat. Kristi; in modern books it is also spelt so, and almost always in hymns and rhymes, ancient as well as modern, e.g. Stríðsmenn þá höfðu krossfest Krist | skiptu í staði fjóra fyrst, Pass. 36. I, 19.1, 3, 10.1, 14.1, 15.2, 16.1, 49.4; Postula kjöri Kristur þrjá, 41; Stríðsmenn Krist úr kúpu færðu, 30.1; Framandi maðr mætti Kristi | hér má fínna hvern það lystir, 30.6, 46.12. Icel. also spell Kristinn, Kristilegr, Christian; kristna, to christen, etc.

β. in the middle of syllables k for c is also used in words of foreign origin, Páskar = Pascha, Passover; dreki = draco; leikmenn = laici; Sikley or Sikiley = Sicilia; Gríkland or Grikkland = Greece. In modern books of the last fifty years ck is turned into kk; and even C in proper names is rendered by K, except where it is sounded as S; thus Icel. spell Caesar, Cicero, Cyprus; for Sesar, Sisero, Syprus, Silisia—although even this may be seen in print of the last ten or twenty years—is a strange novelty. There is but one exception, viz. the proper name Cecilia, which, ever since the Reformation, has been spelt and pronounced Sesselja; where, however, the name occurs in old writers, e.g. the Sturl. i. 52 C, it is always spelt in the Latin form. Latin and foreign words are spelt with c in some MSS. communis-bók, f. a missal, Vm. 52. concurrentis-öld, f. dies concurrentes, Rb. crucis-messa = kross-messa, K.Þ.K.

A digraph ch = k is at times found in MSS., as michill = mikill, etc. C is used in nearly all MSS. to mark 100; the Arabian figures, however, occur for the first time in the Hauks-bók and the chief MSS. of the Njála (all of them MSS. of from the end of the 13th to the beginning of the 14th century), but were again disused till about the time of the Reformation, when they came into use along with print. An inverted c (ɔ̄) is sometimes in very early MSS. used as an abbreviation for con (kon), thus ɔ̄ugr = konungr, ɔ̄a = kona, ɔ̄or = konor = konur; hence the curious blunder in the old Ed. of Páls. S., Bs. i. 140, viz. that a bishop had to take charge of women and clergy instead of choir and clergy, the word cór of the MSS. being mistaken for ɔ̄or (konor). In MSS. of the 15th century c above the line is used as an abbreviation, e.g. tca = taka, tcr = tekr, mcill = mikill, etc.


D (dé) is the fourth letter of the alphabet; it is also written Ð ð (eð). The Gothic Runes have a special sign for the d, or Rounded D rune, namely, a double D turned together; this d is found on the Runic stone at Tune, the Golden horn, and the Bracteats. The reason why this character was used seems to have been that the Latin d Angular Latin D was already employed to mark the th sound (), which does not exist in Latin. The Anglo-Saxon Runes follow the Gothic; again, the common Scandinavian Runes have no d, but use the tenuis t, to mark both d and t.

A. Pronunciation, etc.—The Icel. has a double d sound, one hard (d) and one soft (ð commonly called ‘stungið (cut) dé’); the hard d is sounded as the Engl. d in dale, day, dim, dark; the soft ð as the soft Engl. th in father, mother, brother, but is only used as a final or medial, though it occurs now and then in early MSS. to mark this sound at the beginning of words, e.g. ðar, ðinn, ðegar, but very rarely.

B. Spelling.—In very early Icel. MSS. the soft d in the middle or end of words was represented by þ (th); thus we read, bloþ, faþir, moþir, guþ, orþ, eymþ, sekþ, dypþ, etc., blood… depth, etc. Even Thorodd does not know the form ð, which was borrowed from the A.S. at the end of the 12th century, and was made to serve for the soft th sound in the middle or end of words, þ being only used at the beginning of syllables; but the old spelling with þ in the middle and at the end of syllables long struggled against the Anglo-Saxon ð, and most old vellum MSS. use ð and þ indiscriminately (bloþ and bloð); some use þ as a rule, e.g. Cod. Upsal. (Ub.) of the Edda, written about a.d. 1300, Ed. Arna-Magn. ii. 250 sqq. At the beginning of the 14th century ð prevailed, but again lost its sway, and gave place to d, which marks both the hard and soft d sound in all MSS. from about a.d. 1350 sqq. Thenceforward ð was unknown in Icel. print or writing till it was resumed in the Ed. of Njála a.d. 1772 (cp. also the introduction to the Syntagma de Baptismo, a.d. 1770), and was finally introduced by Rask in common Icel. writing about the beginning of this century; yet many old people still keep on writing d throughout (fadir, modir). On the other hand, Norse (Norwegian) MSS. (laws) never use a middle or final þ; and such words as oþr, goþr in a MS. are a sure mark of its Icel. origin.

C. Changes:

I. assimilation:

1. ðd change into dd, as in the feminines breidd, vídd, sídd, from breiðr, víðr, síðr; pret. beiddi, leiddi, ræddi, hæddi, hlýddi, etc., from beiða, ræða, hlýða, etc.

2. ðt into tt, adj. neut., gott, ótt, brátt, leitt, from góðr, óðr, bráðr, leiðr.

3. the Goth. zd, Germ. and Engl. rd into dd in words such as rödd = Goth. razda; oddr = Germ. ort; hodd = Engl. hoard, Goth. huzd; gaddr = Goth. gazds, etc. Those words, however, are few in number.

II. the initial þ of a pronoun, if suffixed to the verb, changes into ð or d, and even t, e.g. far-ðu, gör-ðu, sjá-ðu, fá-ðu, bú-ðu, = far þú (imperat.), etc.; kalla-ðu, tala-ðu, = kalla þú, tala þú; or kom-du, leid-du, bíd-du, sýn-du, sen-du, = kom þú, leið þú, etc.; or t, hal-tu, vil-tu, skal-tu, ben-tu, hljót-tu, = hald þú, vilt þú, skalt þú, bend þú, hljót þu; and even so the plur. or dual—komi-ðið, haldi-ðið, ætli-ðið, vilið-ið, göri-ðér, gangi-ðér, = komi þið … gangi þér; or following conjunctions, efað-ðú = ef að þú, síðan-ðú = síðan þú, áðren-ðú = áðr en þú.

III. change of d into ð:

1. d, whether radical or inflexive, is spelt and pronounced ð after a vowel and an r or f, g, e.g. blóð, þjóð, biðja, leið, nauð, hæð, brúðr, bæði, borð, orð, garðr, ferð, görð, bragð, lagði, hægð, hafði, höfðum (capitibus), etc. This is without regard to etymology, e.g. Goth. þiuda (gens) and þjuþ (bonum) are equally pronounced and spelt ‘þjóð;’ Goth. dauþs and dêds, Icel. dauði and dáð; Goth. guþ (deus) and gôds (bonus), Icel. guð, góðr; Goth. fadar, bruþar, Icel. faðir, bróðir, cp. Germ. vater, mutter, but bruder; Goth. vaurd and gards, Icel. orð, garðr; Engl. burden and birth, Icel. byrðr, burðr, etc. Again, in some parts of western Icel. rð, gð, and fd are pronounced as rd, gd, fd, ord, Sigurd, gerdu (fac), bragd (with a soft g, but hard d), hafdi (with a soft f and hard d); marks of this may be found in old MSS., e.g. Cod. Reg. (Kb.) of Sæm. Edda.

2. an inflexive d is sounded and spelt ð:

α. after k, p, e.g. in pret. of verbs, steypði, gleypði, klípði, drúpði, gapði, glapði, steikði, ríkði, sekði, hrökði, hneykði, blekði, vakði, blakði, etc., from steypa, klípa, drúpa, gapa, glepja, steikja, ríkja, sekja, hrökkva, hneykja, blekkja, vekja, or vaka, etc.; and feminines, sekð, eykð, dýpð, etc.

β. after the liquids l, m, n in analogous cases, valði, dulði, hulði, deilði, and dæmði, sæmði, dreymði, geymði, samði, framði, and vanði, brenði, etc., from dylja, deila, dreyma, semja, venja, brenna, etc.; feminines or nouns, sæmð, fremð, vanði (use), ynði (delight), anði (breath), synð (sin): these forms are used constantly in very old MSS. (12th century, and into the 13th); but then they changed—lð, mð, nð into ld, md, nd, and kð, pð into kt, pt, etc.

γ. after s (only on Runic stones; even the earliest Icel. MSS. spell st), e.g. raisþi = reisti from reisa. In MSS. of the middle of that century, such as the Ó.H., Cod. Reg. of the Eddas and Grágás, the old forms are still the rule, but the modern occur now and then; the Grágás in nineteen cases out of twenty spells sekð (culpa), but at times also ‘sekt;’ kð, pð were first abolished; the liquids kept the soft d till the end of the century, and lð, mð, nð is still the rule in the Hauksbók; though even the chief vellum MS. of the Njála (Arna-Magn. no. 468) almost constantly uses the modern ld, md, nd. As to kt and pt, the case is peculiar; in early times the Icel. pronounced dýpð or dýpþ etc. exactly as the English at present pronounce depth; but as the Icel. does not allow the concurrence of two different tenues, the modern pt and kt are only addressed to the eye; in fact, when ð became t, the p and k were at once changed into f and g. The Icel. at present says dýft, segt, just as he spells September, October, but is forced to pronounce ‘Seft-,’ ‘Ogt-.’ The spelling in old MSS. gives sometimes a clear evidence as to the etymology of some contested words, e.g. the spelling eykð (q.v.) clearly shews that the word is not akin to Lat. octo, but is derived from auka (augere), because else it would have been formed like nótt, átta, dóttir, Lat. noct-, octo, Gr. θυγάτηρ; so anði, synð, shew that the d in both cases is inflexive, not radical, and that an, syn are the roots, cp. Gr. ἄνεμος and Germ. sühnen; but when editors or transcribers of Icel. MSS.—and even patriotic imitators of the old style—have extended the ð to radical ld, nd, and write lanð, banð, hönð, valð, etc., they go too far and trespass against the law of the language. It is true that ‘land’ is in Icel. MSS. spelt ‘lð,’ but the stroke is a mark of abbreviation, not of a soft d.

D. Interchange (vide p. 49):

I. between Greek, Latin, and Scandinavian there are but few words to record, θυγάτηρ = dóttir, θήρ = dýr, θύρα = dyrr, θάνατος and θνήσκω = dá and deyja, θεός = díar, θαλλός = dalr (arcus), and perhaps θύω = dómr; Lat. truncus = draugr, trabere = draga.

II. between High German on the one hand, and Low German with Scandinavian on the other hand, a regular interchange has taken place analogous to that between Latin-Greek and Teutonic; viz. Scandin.-Engl. d, t, þ answer to H.G. t, z, d, e.g. Icel. dagr, Engl. day = H.G. tag; Icel. temja, Engl. tame = H.G. zähnen; Icel. þing, Engl. thing = H.G. ding.

In very early Icel. MSS. we find the old Latin form d, which sometimes occurs in the Kb. of the Sæm. Edda, but it is commonly , whence ð is formed by putting a stroke on the upper part, A.S. ð; this shews that the ð is in form a d, not a þ (th); vide more on this subject in the introduction to þ: Thorodd calls the capital D edd, the d dé.


E (a), the fifth letter, is in the old Gothic and Anglo-Saxon Runes represented by , being in Anglo-Saxon called ‘eoh;’ the common Scandinavian Runes have no character for e, but mark it either ia or i, and, still later, , with a knob in the middle (‘stunginn Íss’ ).

A. Pronunciation, etc.—The Icel. e is sounded as English a in same, take, and in modern printed books is only used in radical syllables without regard to etymology; but there is sufficient evidence that in early times in Icel. the e had a double sound, one long, like the Italian e or English a (long), the other short, like e in English wet. These two sounds are etymologically different; the first is of comparatively late growth and derived from a by vowel change or otherwise; it is therefore in kindred languages (Swed., Germ.) often spelt ä, so as to indicate its origin from the mother-letter a: the other e is much older, nearly akin to i, being related to that letter as o to u. Grimm suggests that e is derived from i as o from u (only admitting a, i, u as primitive vowels), but in the Icel. at least e and o are in spelling as old as i or u, and seem to be primitive. The Runes in Tune and on the Golden horn have special marks for e and o. At the time of Ari and Thorodd the two seem to have been distinguished in Icel. The latter grammarian uses a special sign for each; he proposes to represent the long sound (Engl. a) by  (commonly ę), adding (as he says) the bight of a to the body of e, to express a sound intermediate between ä and e; he therefore would have written tk (I take), vnja, tmja (to tame), but eðr, en, ef, etc., Skálda 161–163; in the unique vellum MS. (and in Edd.) the characters are not given correctly, as transcriber and editors did not fully understand the bearing of the author’s words. About 700 years later, Jacob Grimm (without knowing the Icel. grammarian or the spelling of MSS. not then edited) recalled the old double e sound to life, guided by the analogy of other Teutonic languages. He proposed to represent a (the ę of Thorodd) by e, and the genuine e by ë. He (Gram. i. 281–284) drew out a list of words founded on the supposed etymology, and kept this distinction wherever he spelt Icel. words. It is curious to observe the difference between Grimm’s artificial list of words and the phonetic spelling in some MSS.; there are especially two MSS., both of them Norse, which are remarkable for their distinction of the two sounds, the long e being spelt with æ, the short with e: these MSS. are the O.H.L., published from a vellum MS. Ups. De la Gard. no. 8, written in Norway at the beginning of the 13th century, and edited by C.R. Unger; the second, small fragments of Norse law MSS., published in N.G.L. ii. 501–515 and i. 339 sqq. Some words compiled from them are as follow:

I. æ: the verbs, bærja, blækkja, ærja, æggja, færja, hængja, glæðja, hæfja, hærja (to harry), kvæðja, læggja, sægja, sælja, sætja, strængja, væðja (to bail), værja, etc.; bænda, brænna (brændi), bræsta, æfla, æfna (Swed. ämna), fælla (to fell), frægna, gægna, hæmna (= hæfna), hværfa (to turn), kænna, mætta, næmna (Swed. nämna), rænna (to let run), ræfsa, spænna, stæmna (stafn), tælja, værða (to become), værka, vækra (vakr), þværra: nouns, bæn, a wound (but ben, N.G.L. iii. 388); bær, a berry; bæðr, a bed; bælgr; bærsærkr; bælti, a belt; dæpill; drængr, a man; drægg; ækkja, a widow; ændi, end; ældr, fire; æmni (= æfni = Swed. ämna); æmbætti (Germ. amt); ældri (in for-ældri, forefathers, Germ. ältern); ælja, a concubine; ærendi, an errand; ærændr, exanimis; ængill, an angel; ærmr, a sleeve (armr); ærvi, ærfingi, ærfð (arfr); ænni, the forehead; ærtog (a coin); æng, a meadow (ang = a sweet smell); Ærlingr (a pr. name); ærki-, Engl. arch- (ἀρχι-); ærveði, toil, and ærveðr, toilsome; ægg, an edge; fæðgar (faðir); fælmtr (falma); færð (fara); frælsi (frjals); hæl, hell; hælviti; hælla, a stone; hællir, a cave; hærra, a lord; hærr, troops; hærbúðir; hærnaðr; hærað, a county (but herað in N.G.L. i. 344 sqq.); hærðar, shoulders; kæfli (Swed. kafle); kær, a jar; kælda (kaldr), a well; kætill, a kettle; fætill; kvæld, evening; kværk, the throat; læggr, a leg; mærki, a mark; mærgð (margr); mægn, mægin, main; mærr, a mare; næf, nose; næss, a ness; ræfill, tapestry; rækkja, a bed; sækt, sake; skægg, beard; skællibrögð; skæpna, a creature (skapa, Dan. skæbne); sværð, a sword; sænna, sound; væfr, weaving; værk (but verk better, N.G.L. i. 339 sqq., cp. virkr): væstr, the west; væl, a trick; vætr, the winter (but vittr or vitr better, N.G.L. ii. 509); vær (in sel-vær); værðr, a meal; þængill, a king; þækja, thatch; þægn, thane; Ængland, England; Ænskr, English; Ænglændingar, the English (Angli); Tæmps, the Thames, etc.: datives, dægi, hændi, vændi, vælli, hætti (höttr), bælki (balkr): adjectives, compar. and superl., fræmri, fræmstr; skæmri, skæmstr; ældri, ælztr; længri, længstr; bætri, bæztr; værri, værstr; hældri, hælztr: sækr, guilty; værðr, due; fæginn; hælgr, holy; bærr, bare; stærkr, stark, etc.: prepositions, hænni, hænnar (hann); tvæggja, duorum; hværr, who; ænginn, none; ækki, nothing (but also engi, which is better), etc.: particles, æftir, after; væl, well; ælligar, or: inflexive syllables, -sæmd (-sanir); -ændi; -spæki, wisdom, etc.: the diphthongs æi and æy = ei and ey, læita, bæita, hæyra, æyra, etc.

II. e: the pronouns and particles, eða, or; ek, ego; enn, still; en, but; sem, which; ef, if; með, with; meðan, while; meðal, between; nema, nisi; snemma, early; er, is, and eru, are; em, I am; þessi, this; þetta, that; sex, six; sek, mek, þek, sometimes instead of sik, mik, þik: nouns, elgr, an elk; sef, sib; brekka, brink; veðr, weather; nevi, a kinsman (Lat. nepos); nevi, a neave, fist; segl, a sail (cp. segla); vetr, a wight; selr, a seal; net, a net; nes, a ness; el, a gale; messa, a mass (Lat. missa); hestr, a horse; prestr, a priest; þegn (O.H.L. 47); vegr, a way, honour; sel and setr, shielings; veröld, the world; vesöld, misery: verbs, gera, to ‘gar,’ to do; drepa, to kill; bera, to bear; bresta, to burst; gefa, to give; geta, to get; meta, to measure; kveða, to say; drekka, to drink; stela, to steal; vera, to be; mega, must; nema, to take; eta, to eat; vega, to weigh; reka, to drive; skera, to cut: participles and supines from þiggja, liggja, biðja, sitja, þegit, legit, beðit, setið: preterites as, hengu, gengu, fengu (Germ. gingen, fingen); greru, reru, sneru (from gróa, róa, snúa): e if sounded as é, e.g. hét, blés, lét, réttr, léttr; even in the words, hér, here; mér, sér, þér, mihi, sibi, tibi; neðan (niðr), hegat (= huc); héðan, hence: adjectives, mestr, flestr, þrennr, etc.: inflexions, -legr, -ly; -lega, -ly; -neskja, -neskr (cp. Germ. -isch); in the articles or the verbal inflexions, -en, -et, -er, -esk, etc. The e is often used against the etymology, as dreki, dragon; menn, men (from maðr). In some other Norse MSS. the two sounds are marked, but so inaccurately that they are almost useless, e.g. the chief MS. of the Barl. S.; but in other MSS. there is hardly an attempt at distinction. The list above is mainly but not strictly in accordance with the etymology, as phonetical peculiarities come in; yet the etymology is the groundwork, modified by the final consonants: both old spelling and modern pronunciation are of value in finding a word’s etymology, e.g. the spelling drængr indicates that it comes from drangr; hærað and hær, troops (but her, here), shew that hærað (hérað) is to be derived from hærr (herr), exercitus, and not from her (hér), etc. The Icel. idiom soon lost the short e sound in radical syllables, and the long e sound (like the Italian e) prevailed throughout; there was then no more need for two signs, and e, prevailed, without regard to etymology. Some few MSS., however, are curious for using æ almost throughout in radical syllables, and thus distinguish between the e in roots and the e in inflexions (vide B below); as an example see the Arna-Magn. no. 748, containing an abridgement of the Edda and Skálda and poems published in the edition of 1852, vol. ii. pp. 397–494; cp. also Vegtamskviða, published by Möbius in Sæm. Edda, pp. 255, 256, from the same MS.; this MS. uses æ in radical syllables, but e or i in inflexions. It is clear that when this MS. was written (at the latter part of the 13th century) the Icel. pronunciation was already the same as at present. In some other MSS. e and æ, and e and ę now and then appear mixed up, till at last the thing was settled in accordance with the living tongue, so that the spelling and sound went on together, and æ (or ę) was only used to mark the diphthong; vide introduction to Æ.

B. Spelling of e and i in inflexions.—The Germans, Swedes, Danes, English, and Dutch all express the i sound in inflexional syllables by e, not i, as in Engl. father, mother, brother, taken, bidden, hidden, heaven, kettle; or in Germ., e.g. hatte, möchte, sollte, lange, bruder, mutter, soltest, himmel, etc.: in the earliest times of Icel. literature also it is almost certain that e was used throughout: Ari probably signed his name Are (en ek heiter Are, Íb. fine): Thorodd, too, seems to have followed the same rule, as we may infer from several things in his treatise, e.g. the words framer and frá mér, which would be unintelligible unless we suppose him to have written framer, not framir: even the name of Snorri is twice spelt Snorre in the Reykholts-máldagi, probably written by one of his clerks. Some old vellum fragments may be found with the e only; but even in the oldest extant, i is used now and then. The reason is clear, viz. that the Icel. never admits the long e in inflexive syllables, and in roots it never admits the short e, consequently the same sign would not do both for roots and inflexions; hende, velle, gefe have each two vowel sounds; therefore the short i was admitted in inflexions; yet in most MSS. both e and i are used indiscriminately, e.g. faðir and faðer, tími and time, manni and manne, kominn and komenn, komið and komet, höndin and hönden, fjallit and fjallet; even those that use i admit e if following ð or d, e.g. viðe, bæðe, liðe, lande, but fjalli, vatni. As the spelling was partly influenced from abroad, the e even gained ground, and at the time of the Reformation, when printing became common, it was reassumed throughout, and remained so for nearly 230 years, when (about a.d. 1770–1780) i was reinstated and e expelled in all inflexions, as being inconsistent with the spelling and ambiguous; but the sound has undoubtedly remained unchanged from the time of Ari up to the present time: the English father, mother, German vater, mutter, and Icel. fadir are, as to the inflexion, sounded exactly alike.

C. Interchange of e and i.—The adjectival syllable -ligr, -liga, is in MSS. spelt either -ligr or -legr; in modern pronunciation and spelling always -legr, -lega (Engl. -ly).

β. in a few root words e has taken the place of i, as in verðr, qs. virðr (food); brenna, qs. brinna; þremr and þrimr; tvenna and tvinna; ef, efa, efi, = if, ifa, ifi; einbirni and einberni (born): e has taken the place of ø in such words as hnetr (nuts) from hnot, older form hnøtr: so also in eðli and öðli; efri efstr from öfri öfstr: e and the derived ja make different words, as berg and bjarg, fell and fjall, bergr and bjargar, etc.

D. Diphthongs:

I. ei answers to Goth. ai, A.S. â, Germ. ei, Engl. ō (oa or the like); in Danish frequently expressed by ee; in Swedish and Northern English the diphthong is turned into a plain e and a, which, however, represent the same sound: Goth. stains, A.S. stân, Swed. sten, North. E. stane. The o sound is English-Saxon; the a sound English-Scandinavian; thus the forms, home, bone, oak, oath, broad, one, own, more, none, no, may be called English-Saxon, from A.S. hâm, bân, etc.; the North. E. and Scottish hame, bane, aik, aith, braid, ain, mair, nain, may be called English-Scandinavian: cp. Swed. hem, ben, ek, ed, bred, en; Icel. heimr, bein, eik, eiðr, breiðr, einn, meir, neinn, nei; cp. also Icel. bleikr, Swed. blek, North. E. blake, etc. The Runic stones mark the ei with a + i or i simply, e.g. stin or stain. Old Norse and Icel. MSS. frequently for ei give æi.

II. ey is in modern usage sounded as ei, and only distinguished in writing; in old times a distinction was made in sound between ei and ey. Norse MSS. almost always spell öy, and in Norway it is to the present time sounded accordingly, e.g. öyra, = Icel. eyra, sounded nearly as in English toil: the ey is properly a vowel change of au: ey frequently answers to an English e (ea) sound, as heyra, to hear; eyra, ear; dreyma, to dream; leysa, to lease. In very old MSS., e.g. Íb. (ai in the Ed. is a wrong reading from ꜹ in the MS.), au and ey are even spelt alike (ꜹ or au), though sounded differently. In some MSS. ey is also used where it is not etymological, viz. instead of ø or ö, in such words as hreyqva, seyqva, steyqva, deyqvan, greyri, geyra, seyni, etc., = hrökva, sökva, … greri or grori, syni, e.g. the Cod. Reg. of Sæm. Edda, the Rafns S. Bs. i. 639 sqq.

E. é is sounded almost as English ye (or ya); it is produced,

1. by an absorption of consonants, in words as réttr, léttr, þéttr, sétti, flétta, rétta, cp. Germ. recht, Engl. right; Germ. leicht, Engl. light: or in fé, kné, tré, hlé, sé (Icel. fé = Engl. fee, Goth. faihu, Lat. pecus), etc.

2. by a lost reduplication in the preterites, féll, grét, réð, lét, blés, hét, gékk, hékk, lék, fékk, from falla, gráta, etc.; in some old MSS. this é is replaced by ie, e.g. in the Hulda Arna-Magn. no. 66 fol. we read fiell, liet, hiet, griet, gieck, liek, cp. mod. Geim. fiel, hiess, liess, etc.; perhaps in these cases é was sounded a little differently, almost as a bisyllable.

3. in such words as the pronouns vér, þér or ér (you), mér, sér, þér (tibi): the particles hér (here), héðan (hence), hérað, vél, él.

4. é is also sounded after g and k, and often spelt ie in MSS., gieta, giefa, kier, kierti; this sound is, however, better attributed to g and k being aspirate. In Thorodd and the earliest MSS. é is marked with ´ just like the other long or diphthongal vowels; but the accent was subsequently removed, and e and é are undistinguished in most MSS.: again, in the 15th century transcribers began to write ie or ee (mier or meer). In printed books up to about 1770 the ie prevailed, then e, and lastly (about 1786) é (cp. the 5th and 6th vols. of Fél.): è is an innovation of Rask, and is used by many, but máttr, dráttr, and réttr, sléttr, etc. are etymologically identical, though the sound of é is somewhat peculiar: the spelling je is also a novelty, and being etymologically wrong (except in 2 above) is not to be recommended.


F (eff), the sixth letter, was in the Gothic Runes, on the Bracteats, and on the stone in Tune, marked , a form evidently derived from the Greek and Latin; hence also comes the Anglo-Saxon called feoh, and in the Scandinavian Runes Curved Runic F called (= fee, money), fé veldr frænda rógi, Rkv. I. The Runic alphabet makes f the first letter, whence this alphabet is sometimes by modern writers called Fuþork. The first six letters are called Freys-ætt, the family of Frey; perhaps the Goths called this Rune Frauja = Freyr, the lord. Only in very early Icel. MSS. is the old Latin form of f used: at the beginning of the 13th century the Anglo-Saxon form Anglo-Saxon F (derived from the Rune) prevailed; and it was employed in printed Icel. books till about a.d. 1770, when the Latin f came into use. In very early MSS. ff and ft are very difficult to distinguish from ſf and ſt. Emendations may sometimes be made by bearing this in mind, e.g. hóstú, Am. 95, should clearly be read hóftú = hóft þú, from hefja,—proving that this poem was in writing not later than about a.d. 1200, when the Anglo-Saxon letter was introduced.

A. Pronunciation.—At the beginning of a syllable always sounded as Engl. f; but as a medial and final, it is often pronounced and sometimes spelt v, especially after a vowel, so that in af, ef, lauf, gefa, hafa, grafa, lifa, líf, gröf, f is pronounced like the v, as in Engl. grave. Foreign proper names, Stefan (Stephen), etc., are exceptions, where f not initial has an aspirate sound. For the exceptional spelling of f as b vide introduction to B, (pp. 48, 49.) The Icel. dislike a double f sound, which is only found in a very few modern foreign words, such as kaffe, coffee; straff, Germ. strafe, punishment; koffort, a box (from French or Germ.); offur, an offer; skoffin, a monster; skeffa, a ‘skep’ or bushel; skúffa, a drawer; eff, the name of the letter itself, cp. Skálda 166.

B. Spelling:

I. as an initial the spelling never changes; as medial and final the form f is usually retained, as in álfr, kálfr, sjálfr, silfr, arfr, orf, úlfr, etc., af, gaf, haf, etc., although the sound is soft in all these syllables. Some MSS. used to spell fu, especially after an l, sialfuan (ipsum), halfuan (dimidium), etc.; in the 14th century this was common, but did not continue; in Swedish it prevailed, hence the mod. Swed. forms gifva, drifva, etc.

II. the spelling with f is against the true etymology in many cases, and here also the spelling differs; this is especially the case with the final radical v or u (after a vowel or after l or r), which, being in some cases suppressed or obsolete, reappears and is differently spelt; thus, örfar, arrows (from ör); snjófar (nives), snow, and snjófa, to snow (from snjór); háfan (acc.), high (from hár); mjófan, thin (from mjór); sæfar (gen.), the sea (from sær): the partly obsolete dat. forms ölvi, mjölvi, Mávi, bölvi, heyvi, hörvi, smjörvi, lævi from öl (ale), mjöl (meal), hey (hay), etc. are also spelt ölfi … heyfi, cp. e.g. Eb. 94 new Ed. note 8: so also adjectives, as örfan (acc. from örr), liberal: nouns, as völfa or völva, a prophetess.

III. the spelling with pt in such words as, aptan, evening; aptr, after; leiptr, lightning; dript, drift; dupt, Germ. duft; heipt, cp. Germ. heftig; kraptr, Germ. kraft;; aptari, eptri, = aftari, eftri, aft, behind; eptir, after; skipta, to shift; lopt, Germ. luft; kjöptr, Germ. kiefe; opt, often; nipt (from nefi), a sister; hapt, a haft, hepta, to haft; gipta, a gift; raptr, a rafter; tópt, cp. Engl. toft, Dan. toft; skapt, Engl. shaft, Dan. skaft; þopta, Dan. tofte,—is against the sense and etymology and is an imitation of Latin MSS. The earliest MSS. and almost all Norse MSS. use ft, and so also many Icel. MSS., e.g. the Flateyjar-bók, Hauks-bók, etc.; pt, however, is the regular spelling, and hence it came into print. The present rule appears to be to use pt wherever both consonants are radicals, but ft if the t be inflexive—thus haft, part. from hafa, lift from lifa, hlíft from hlífa; but in speaking pt and ft are both sounded alike, regardless of etymology, viz. both as ft or vt with a soft f sound; hence phonetic spelling now and then occurs in MSS., e.g. draft = drapt, from drepa, Fb. i. 149; efðe = æpði = æpti, from æpa, to weep, Bs. i. 342; keyfti, from kaupa, Greg. 50; steyfti, from steypa.

β. a digraph fp or pf occurs a few times in MSS., efptir, 673 A. 2; lopfti = lopti, Greg. 72 (vide Frump. 100), but it never came into use; it reminds one of the pf which in modern German is so frequent: fmf or m, e.g. nafm—nafn or namn, Mork. 60 and N.G.L. passim; fft = ft also occurs in old MSS.

C. Changes.—The final soft Icel. f answers to Engl. f, ve, e.g. Icel. líf = Engl. life, but Icel. lifa = Engl. to live; gefa, to give; hafa, to have; leifa, to leave. Again, the spurious Icel. f (B. II) usually answers to Engl. w or the like, e.g. örfar = Engl. arrow; snjófar = Engl. snow; már náfi, cp. Engl. mew; Icel. nær (the v is here suppressed), cp. Engl. narrow; Icel. lævi, cp. Engl. lewd, etc. etc. In Danish the soft f is usually spelt with v, e.g. halv, kalv, hav, give, love, sove,—Icel. hálfr, kálfr, haf, gefa, lofa, sofa, whereas the Swedes frequently keep the f. In German a final b answers to Icel. f; Germ. geben = Icel. gefa, Engl. give; Germ. kalb, erbe, = Icel. kalfr, arfi, etc., see introduction to B. Again, in German a final f or ff answers to Icel. and Engl. p, e.g. Germ. lauf = Icel. hlaup, Engl. leap; Germ. kauf = Icel. kaup, Engl. cheap; Germ. schiff = Icel. skip, Engl. ship, also skiff; Germ. treff = Icel. drep; Germ. tief = Icel. djúpr, Engl. deep; Germ. haufen = Icel. hópr, Engl. heap; Germ. rufen = Icel. hrópa; Germ. schaffen = Icel. skapa, Engl. shape; Germ. saufen = Icel. súpa, Engl. to sup; Germ. hofte = Icel. huppr, Engl. hip; Germ. greifen = Icel. grípa, Engl. to grapple, grip; Germ. gaffen = Icel. gapa, Engl. gape; Germ. offen = Icel. opinn, Engl. open; Germ. affe = Icel. api, Engl. ape; Germ. triefen = Icel. drjúpa, Engl. drip; Germ. tropfen = Icel. dropi, Engl. drop. As to the use of the initial f, the Engl., Icel., Swed., and Dan. all agree; the High Germ. spelling is confused, using either f or v, but both of them are sounded alike, thus voll = Engl. full, Icel. fullr; vier = Engl. four, Icel. fjórir; vater = Engl. father, Icel. faðir, etc.: but fisch = Engl. fish, Icel. fiskr; fest = Engl. fast, Icel. fastr. This German v, however, seems to be dying out (Grimm, introduction to F).

2. for the change of fn and mn, see introduction to B: f changes to m in a few Icel. words, as himin, qs. hiffin, cp. Engl. heaven; helmingr, a half, from halfr, half.

D. Interchange.—The Greek and Latin p answers to Teutonic and Icel. f; thus, pater, paucus, piscis, πέντε, πῦρ, πῶλος, pecu, pellis, πίων, pinguis, plecto, pes, ποδ-ός, pallor, etc., cp. Icel. faðir, fár, fiskr, fimm, furr, foli, fé, fell (feldr), feitr, flétta, fet and fótr, fölr, etc.; Lat. portare = færa, Engl. to ford; se-pelio = fela; πτερόν = fjöðr and fiðr; πνέω and πνεῦμα, cp. fnasa; Lat. per, pro, προ-, cp. fyrir; Lat. plēnus, pleo, πλέον, πλέος, cp. fullr; πλοῖον = fley; Lat. prior, πρῶτος, cp. fyrir, fyrstr: Lat. primus, cp. frum-; Lat. plūres, plērique, πολλός, πλεῖστος, πολύς, = fjöl-, fjöd, fleiri, flestr; Lat. plicare = falda; Lat. prētium, cp. friðr, fríðendi, etc. (vide Grimm). Again, where no interchange has taken place the word is usually borrowed from the Greek or Latin, e.g. forkr, Engl. fork = Lat. furca; Icel. fals, falskr = Lat. falsus; Icel. fálki = Lat. falco, etc.


G (gé) is the seventh letter. In the old Gothic Runic alphabet (Golden horn) it is represented by , which was probably taken from the Greek χ. The later common Runic alphabet had no g, and made the tenuis k (, called Kaun) serve for both; still later, g was distinguished simply by a dot or stroke, or , and this character was called ‘Stunginn Kaun,’ i.e. dotted or cut Kaun, just as the name of Stunginn Týr was given to cut or dotted t.

A. In Scandinavia the letter g begins many fewer words than in German or Saxon, mainly because the prefixed particle ge- is absent. In the fragments of Ulf., although so little is left, ga- is prefixed to about three hundred words, mostly verbs and nouns; in the Anglo-Saxon at least three or four thousand such words are recorded, and in modern German still more: indeed the number is so to say endless, as it can be put to almost any verb. In Icel. the only traces of this prefix are,

I. in a few words retaining g before the liquids l and n (gl and gn):

α. gl in the word glíkr, similis (and derivatives); glíkr is now obsolete, and even in very old MSS. of the 13th or even the 12th century both forms, glíkr and líkr, glíkendi and líkendi, glíkjast and líkjast, occur indiscriminately; but in older poems gl is the only form.

β. gn in gnadd, gnaga, gnauða, gnegg, gneisti, gnípa, gnísta, gnolla, gnógr, gnúa, gnúpr, gnyðr, gnæðingr, gnöllra, gnötra (q.v.), and some poët. words, as gnat, etc. But in mod. usage, in gn and gl, the g is dropped both in spelling and pronunciation, nadd, naga, nauða, hnegg, neisti, nípa … núpr, nyðr or niðr, næðingr, nöllra, nötra; the gn in these words is almost constantly used in very old MSS., but even at the end of the 13th and in the 14th century the MSS., e.g. Hb., begin to drop the g, vide p. 206 sqq.: the exceptions are few, e.g. Icel. never say nýja for gnýja, but the word itself, although known, is almost obsolete: so also in modern writers gnótt and gnægtir (abundance) often occur: but the sound gn may be said to be almost extinct. The Danes, Swedes, and Norse still keep the g before n, e.g. Dan. gnave, Swed. gnaga; whereas in glíkr the g has been dropped, and the word has become in Swed. lik, etc.; in Dan. lig, lige, ligning, etc.

II. in two Icel. words the prefixed g has hardened into a radical consonant, so that its proper sound is no longer perceived, viz. granni (and compds), a neighbour, prop. one of the next house, Goth. garazna = γείτων, qs. g-ranni, from rann, domus; and greiða, explicare, = Goth. garaidian. The Scandinavian tongues have furthermore done away with the Saxon and German prefix to passive participles, and no trace of them remains even in the earliest writers or poems. The modern English has followed the same law as the Scandinavian in gn, for though it still appears in Engl. words (as gnaw, gnash), it is hardly sounded. The participial prefix remained long in southern England (see Morris’s Specimens), but weakened into y or i till at last it dropped altogether.

B. Pronunciation.—It is sounded hard, soft, or aspirate; hard, as in Engl. gate, gold; soft, as in Swed. dag, Germ. tag, or mod. Gr. γ, but lost in Engl.; aspirate also lost in Engl.

I. hard,

1. as initial before a hard vowel, garðr, gull, gott, etc.; and before a consonant, glaðr, gráta; but the prefixed g, in the instances A. 1. above, was prob. always sounded soft.

2. as final after consonants, as sorg, belg, ung, höfgi, or if double, as in egg.

II. soft, never as initial (unlike mod. Greek, in which γ is sounded soft throughout), but only as final or sometimes as medial:

1. if single after a vowel, as dag, hug, log, veg, stig.

2. between two vowels if the latter is hard, lega, ligum, vega, vegum, dögum; but in case both the vowels, or even only the last, are soft (an i vowel) the g sound is lost, and it is eliminated altogether or assimilated to the preceding vowel, which thus becomes a diphthong; the same is the case if j follows g; thus syllables and words such as bagi and bæi, dagi and dæi, degi and deigi, eygja and eyja or eya, lagi and lægi or læi are all sounded alike; in olden times there must have been a difference of sound, as old MSS. never confound the spelling in words like those above, whereas in modern letters written by uneducated people, nothing is more frequent than to see, um dæinn for um daginn, or á deíinum for á deginum, and the like; the poets also rhyme accordingly, e.g. segi—hneigi, Pass. 38. 13; segja—deyja, 25; segja—beygja, 25. 12; drýgja—nýja, 30. 3; eigið—dregið, 7. 10; deyja—teygja, 16. 13, etc.; even MSS. of the end of the 15th century frequently give seigia for segja (to say), e.g. Arna-Magn. 556 A, see the pref. to Ísl. ii. p. vi: as a medial, before d the g is sounded hard almost all over Icel., and the d soft (sagði); yet in the peninsula of Snæfells Sýsla many people still reverse this rule, and say sagdi, lagdi, bygdi, bygd, sounding the g soft but the d hard; in the east of Icel. people say bregða, sagði, pronouncing both soft; this is no doubt the best pronunciation, and accords well with the modern English said, laid, and the like.

III. the aspirate g is sounded,

1. as initial before a soft vowel or j, gefa, gæta, geyma, geir, gjöld.

2. as final, a double g (gg) or g after a consonant is sounded as aspirate in all instances where a single g is lost (vide above), thus laggir, leggja, byggja, byggi, veggir, or margir, helgir, göfgi, engi, mergjar, elgjar, engja. Between two consonants the g is not pronounced, thus fylgdi, morgna, fylgsni, bólgna are sounded as fyldi, morna, fylsni, bólnar.

C. Spelling.—Here is little to notice:

I. in old MSS. the aspirate g as initial is frequently marked by the insertion of i after it, thus giæta, giefa, = gæta, gefa, but this is not now used.

II. in old Norse MSS.,—and, by way of imitation, in some Icel.,—the soft g before a vowel is frequently marked by inserting h after it, thus dagh, deghi, vegha, sagha; in the Middle Ages many foreign MSS. expressed soft sounds in this way, and so they wrote dh = ð, gh = soft g, th = þ, whence comes the th in modern English; we also find gh in words such as Helghi, Fb. pref.; probably the g was in olden times sounded soft in rg, lg, which agrees with the change in English into holy, sorrow, etc.; ngh = ng also occurs, e.g. erlinghi, Fb. i. 537, denoting a soft sound of ng as in modern Danish and Swedish. In MSS. we now and then find a spurious g before j and a vowel, e.g. deygja, meygja, for deyja, meyja, because the sound was the same in both cases.

D. Changes.—The hard and aspirate g, especially as initial, usually remains in modern foreign languages, gate, ghost, give, get, except in Engl. yard, yarn (Icel. garð, garn), etc., where the Anglo-Saxon had a soft g sound. Again,

1. the soft g after a vowel takes a vowel sound, and is in English marked by w, y, or the like, day, say, saw, law, bow, way, low, = Icel. dag, segja, sög, lög, bogi, veg, lág, etc.: and even a double g, as in lay, buy, = Icel. leggja or liggja, byggja.

2. so also before or after a consonant, thus, Engl. said, rain, gain, sail, tail, bail, fowl, etc., = Icel. sagði, regn, gagn, segl, tagl, hagl, fugl; Engl. sorrow, follow, fellow, worry, borrow, belly, = Icel. sorg, fylgja, félagi, vargr, byrgi, belgr. In Dan. lov, skov, vej answer to Icel. lög, skóg, veg, whereas Sweden and Norway have kept the g, Swed. lag, skog, väg.

E. Interchange.—Lat. h and Gr. χ answer to Icel. and Teut. g, but the instances of such interchange are few, e.g. Lat. hostis, hortus, homo, hoedus, heri, = Icel. gestr, garðr, gumi, geit, gær; Lat. hio, Gr. χάος, cp. Icel. gjá, gína; Gr. χθές = gær, χήν = gáss, χολή = gall, etc.


H (há) is the eighth letter. In the old Runic alphabet it was represented by and   ᚺ  , which are used indiscriminately (but never Greek Eta or Greek Eta with extra lines): and   ᚺ   both occur on the Golden horn, the former once, the latter twice. This Rune was no doubt borrowed from the Greek or Latin. In the later common Runic alphabet this character was replaced by (rarely     ), which we may infer was taken from the Greek χ (the g of the old Runic alphabet) marked with a perpendicular stroke down the middle, rather than from the Latin Latin curved H (see Ritschl’s essay in the Rheinisches Museum, 1869, p. 22); yet the old form   ᚺ   is now and then found on the oldest of the later monuments, e.g. the stones from Snoldelöv, Höjetostrup, and Helnæs (Thorsen 15, 17, 335), on which monuments the is used for a: in some inscriptions serves both for h and soft g. The name of the Rune h was Hagall or Hagl, an Anglo-Saxon form, explained as meaning hail, hagl er kaldast korna (hail is the coldest of grains), in the Norse Runic poem; cp. hägl byð hwitust corna in the Anglo-Saxon poem, which is the prototype of the Norse. These names in the Anglo-Saxon and Norse poems are in no way derived from the form of the Rune, but are merely alike to the modern rhymes in English ABC books,—‘B is a Baker’ or the like. The Hagall was the first of the second group of Runes, H n i a s, which was therefore called Hagals-ætt, the family of Hagal (cp. introduction to F).

A. Pronunciation and Spelling.—H is sounded as in English hard, house: the aspirate is still sounded in hl, hr, hn much as in the Welsh ll, rh: the hv is in the west and north of Icel. sounded as kv; but in the south and east the distinction is kept between hv and kv (hver a kettle and kver a quire, hvölum whales and kvölum torments), as also in writing; and hv is sounded like wh in Northern English; in a small part of eastern Icel. it is sounded like Greek χ (hvalr as χalr, hvað as χað), and this is probably the oldest and truest representation of the hv sound.

II. the h is dropped,

1. in the article inn, in, it, for hinn, hin, hit, which is often spelt so in old MSS.

β. in the personal pronoun hann, hún if following after another word, e.g. ef ’ann (ef hann), ef ’ún (ef hún), þó ’onum (þó honum), látt’ ’ann vera (láttu hann vera), segð’ ’enn’ að koma (segðu henni að koma); this is the constant pronunciation of the present time, but in writing the h is kept: whereas, at the beginning of a sentence the h is sounded, e.g. hann (hón) kom, he (she) came, but kom ’ann? (if asking the question).

γ. in a few words such as álfa and hálfa, óst and host (cp. hósta), ökulbrækr and hökulbraekr.

2. in the latter part of such compounds as have nearly become inflexions, as ein-arðr for ein-harðr: in -úð, -ýðgi, -ygð (Gramm. p. xxxiii, col. 1); elsk-ogi, var-ugi, öl-ogi, from hugr; örv-endr, tröll-endr, gram-endr, from hendr; litar-apt = litar-hapt: in -ald = hald, handar-ald, haf-ald; lík-amr = lík-hamr, hár-amr = hár-hamr; skauf-ali, rang-ali, from hali; at-æfi = at-hæfi, and perhaps in auð-œfi, ör-œfi, from hóf or hœfi; and-œfa = and-hœfa, to respond; hnapp-elda = hnapp-helda: in pr. names in -arr, -alli, -eiðr, -ildr, for -harr = herr, -halli, -heiðr, -hildr, (Ein-arr, Þór-alli, Ragn-eiðr, Yngv-ildr, etc.) In a few words, as hjúpr, and derivatives from júpr, hilmr and ilmr, hopa and opa, h seems to have been added. In some of the cases above cited both forms are still heard, but the apocopate are more usual.

III. h is neither written nor sounded as final or medial, and has in all such cases been absorbed by the preceding vowel or simply dropped (see Gramm. p. xxx, col. 1).

IV. some MSS., especially Norse, use a double form gh and th to mark a soft or aspirate sound, e.g. sagha and saga, thing and þing; especially in inflexive syllables, -ith = -it, etc.

V. a curious instance of spelling (as in Welsh) rh for hr is found occasionally in Runes, e.g. Rhruulfr for Hrúlfr, Thorsen 335; to this corresponds the English spelling wh for hw, in white, wheat, whale, where, whence, why, whelp, whine, whet, whirl, wharf, wheel, while, whim, = Icel. hvítr, hveiti, hvalr, hvar, hvaðan, hvé, hválpr, hvína, hvetja, hvirfill, hvarf, hvel, hvíld, hvima, etc.

B. Remarks, Changes, etc.—In Icel. h is used as an initial letter most largely next to s; in modern Teutonic languages it has been greatly reduced through the dropping of the aspiration before the liquids l, n, r, and before v, whereby all words in hl, hn, hr, and hv have been transferred to the liquids and to v (see Gramm. p. xxxvi, signif. II. β); the h in these words is essential to the etymology, and was in olden times common to all Teutonic languages, but in Scandinavia it was lost about the 11th or 12th century, so that not a single instance of hl, hn, hr is on record in any MS. written in Norway; though old Norwegian poets of the 10th century used it in alliteration, so it must have been sounded at that time; h in hl, hn, hr is therefore a test of a MS. being Icelandic and not Norse. In modern Icel. pronunciation the h aspirate has been lost in two or three words, as leiti for hleyti, a part, a word which was borrowed from Norway about the 14th century; rót = hrót, a roof: it is a matter of course that the h is dropped in words which were borrowed from the English not earlier than the 12th century, e.g. lávarðr, Early Engl. lauerd (lord), but A.S. hlâford.

II. the h has been added in a few words to which it does not rightly belong, viz. in hneiss and hneisa for neiss and neisa; hnýsa for nýsa; hreifr (glad) for reifr; hniðra (to lower) for niðra (niðr); hlykkr (and hlykkjóttr), a curve, for lykkr (cp. lykkja, a noose); hrjóta for rjóta, to snort; hlað, a lace, cp. Lat. laqueus; hnestla for nestla, a loop.

β. in a few instances both forms are used to form double words, in hrífa and rífa, Lat. rapere; hrasa and rasa, to stumble; rata (= Goth. vraton), to find the way, but hrata, to fall (cp. Vsp. 52); hrjá and rjá, to vex.

2. the h seems to be a substitute,

α. for an old v, in hrekja, to toss about, to wreck, akin to Goth. vrekan, Icel. reka; in hreiðr, a nest, Dan. rede, cp. Engl. wreath, Goth. vriþus.

β. in modern pronunciation h is a substitute for g in the words hneggja, hnegg, for gneggja, gnegg; þver-hnípt for þver-gnípt.

γ. for k in hnífr, hnúi, hnefi,* hnöttr, hnútr, hnörr, hnakkr, hnjúkr, hnoða, hnappr, for knífr, knúi, knefi …, knoða, knappr; this spelling is found in MSS. of the 15th century, e.g. the Hrokkinskinna passim (see letter K). In all these cases the h is etymologically wrong; in some of the words above (as in hneisa) it is found even in very old MSS., e.g. the Mork.; but the true etymology is seen from the alliteration in old poems, e.g. Hm. 48, 140, Hðm. 26 (raut, reginkunni); Stor. 13 (Nýsumk hins | ok hygg at því); Edda 105 (reifr gékk herr und hlífar | hizig …); but not so in modern poets, e.g. Hröktu því svo og hrjáðu þig | Herra minn ílsku-þjóðir, Pass. 9. 9; Forvitnin holdsins hnýsir þrátt | í Herrans leyndar-dóma, 21. 2; Nær eg fell eðr hrasa hér | hæstur Drottinn vill reiðast mér, 5. 6.

III. the Gothic has a special sign for hv, viz. w, which thus answers to wh in English, e.g. wan = when.

2. when followed by an o or u, the v in hv is dropped, e.g. hót hooting, hóta to hoot, cp. Goth. wota and wotjan; as also in hót = hvat what, hóll from hváll, hjól and hvel, hólf and hválf, horfinn, hurfu, hyrfi for hvorfinn, hvurfu, hvyrfi.

C. Interchange.—Latin c and Greek κ answer to the Teut. and Icel. h; thus Lat. căper, căput, cănis, carbasus, centum, cervus, cŏr (cord-), collum, corvus, cūtis, = Icel. hafr, höfuð, hundr, hörr (hörv-), hundrað, hjörtr (hirtu) and hjarta, háls (hals), hrafn, húð; calx, cp. hæll; cardo, cp. hjarri; claudus, cp. haltr; clīvus, cp. hlíð; corpus, cp. hræ (hræv-); cĕrebrum, cp. hjarni; crāter, cp. hurð; cōs, cp. hein; clāmo, cp. hljómr; cēlo, cp. hylja and Hel; coelum, cp. holr (hollow); căpio (-cĭpio) = hefja; prin-cipium = upp-haf; cēteri, cp. hindri; co- and con-, cp. hjá; cĭtra, cp. héðra (hér is a contracted form); clūnis, cp. hlaun; clīno, cp. hlein, Engl. to lean; căleo, cp. hlé-, hlý-r; cŏlo, cp. halda; custodio, cp. hodd, Engl. to hoard; cella, cp. hellir; carcer, cp. hörgr; circus, cp. hringr; cŏrium, cp. hörund; curvus, cp. hverfa (to turn round): Gr. καλλίων, κάλλιστος = Icel. hellri, hellztr (hölztr); κᾰλαμος, κεφᾰλή, κέρας, κύων, καρδία, = Icel. hálmr, höfuð, horn, hundr, hjarta; κῶνος, cp. húnn; κλῆρος, cp. hlutr; κυκλός, cp. högl-d, hvel, hjól; κοῖλος, cp. holr; κόραξ, cp. hrafn; κρέας, cp. hræ; κρανίον, cp. hjarni and hvern or hvörn (the two pebble-like bones in a fish’s head), cp. also Goth. wairnis; κρῑός, cp. Icel. hrútr; κρᾰτος, κρᾰτερός, cp. harðr, hraustr; κείρω, cp. herja; κᾰλύπτω, cp. hylja; κλίνω, cp. hlín, hlein; κλύω, cp. hlýða; κρίζω, cp. hrikta; κρηπίς, cp. hriflingar, hrifla; κώμη, cp. heimr; κῦμα, cp. húm; κοινός, cp. hjú-, hjú-n: Lat. quis = hverr; qui = hve; quies, cp. hvíl-d, etc.: some of these words may be dubious, but others are evident.


I is the ninth letter; in the old runic alphabet it was called íss or ice (skálda 176), and represented by (ís köllum brú breiða of the runic poem), a form borrowed from the greek or latin: but ‘stunginn íss’ () was in later runes used to represent e.

A. pronunciation, spelling.—i is either a vowel (i), or consonant (j), called joð: these are here treated separately:

1. the vowel i is sounded either short (i) or long (í), the short (i) like engl. hill, prolonged with a breath; but it is almost certain that in olden times it was sounded short, as in engl. wit.

2. the long (í) is sounded as engl. e or ee in evil, feet.

3. the j is sounded as engl. y before a vowel, jata, jarð, jól, as yata, yard, yole. the oldest writers bear witness to the use of j as a consonant; thus Thorodd says,—i þá er hann verðr fyrir samhljóðanda settr, skálda 164; and the second grammarian,—en ef hljóðstafr (vowel) er næstr eptir hann, þá skiptisk hann í málstaf (consonant), svo sem já, jörð eða jór, 170; and Olave Hvítaskáld,—i ok u hafa því fleiri greinir, at þeir eru stundum samhljóðendr, sem í þessum orðum, iarl and uitr, 176; but in syllables beginning with j (ja, jo, ju) in old alliterative poetry it always stands for the vowel, from the earliest poems down to the 15th century, e.g. jörð or ægi—iðja-græna, Vsp. 58; viltú nokkut jötuninn eiga | ýtum görir hann kosti seiga, þrymlur 2. 2; ölmóðr hafði annan dag | járnið þetta at sýna, skíða r. 64, which, as now pronounced, would sound harsh, since in modern poetry syllables beginning with j cannot be used alliteratively with any other letter, cp. pass. 37. 1, 10, 40. 8, 46. 3, 11, etc.; only in such words as eg (jeg), eta (jeta) can i serve both as a vowel and consonant, see pass. 6. 2; but jeg in 5. 5, 10, (the verse 6 of the same hymn is a poetical licence); so also the name jesús is now and then used alliteratively with a vowel, 47, 18, 21; the hymns of the Reformation follow the same usage. The pronunciation of j seems therefore to have changed: in early times it was probably similar to Engl. e in ear, tear, hear; an additional proof of this is, that the oldest spelling was, as in Anglo-Saxon, ea, eo …; and Thorodd himself probably wrote ea, e.g. eafn, eárn, earl, for jafn, járn, jarl, see his words: in old poets ea sometimes makes two syllables, e.g. in the verse cited in Skálda 164 (of a.d. 1018); as also in the name Njáll (Niel), which is dissyllabic in the verses, Nj. ch. 136, 146. At a still earlier time j was probably sounded purely as a vowel.

II. in ancient MSS. i serves for both i and j; in MSS., esp. of the 15th century, j is used ornamentally for initial i, e.g. jnn = inn, as also in the double ij = í, e.g. tijd = tíð, mijtt = mítt, the j was introduced into print only in the last year of the eighteenth century.

2. an i is often inserted in MSS., esp. after g, k, so as to mark the aspirate sound, e.g. gieta = geta, giæta = gæta, kiær = kær, etc.: in inflexions it is also more correct to write eyjar, bæjar, than eyar, bæar:—ji is not written, but pronounced, e.g. vili (= vilji), but vilja.

B. Changes.—The i and e are exchanged in many root syllables, but i is usually the older, e the later if not the modern form, as, if and ef, brinna and brenna, tvinnr and tvennr, þrimr and þremr, miðil and meðal, snimma and snemma, gingu and gengu, fingu and fengu, tigr and tegr: the article varies between enn and inn:—the inflex. -endi and -indi:—Norse MSS. spell mek, þek, sek, = mik, þik, sik (e.g. Thom. Cd. Holm.); -ligr and -legr, gagnligr and gagnlegr: for the inflexive e and i see introduction to letter E (signif. B), p. 114:—i for y in old MSS., in firir, ifir, mindi, skildi, minni (mouth), minnast (to kiss, mouth):—i and u are interchanged in inflexion, as, morginn and morgunn, vandill and vöndull; but esp. in the adjective inflexions -igr and -ugr, blóðigr and blóðugr, auðigr and auðugr.

II. the j in most instances originates from an e, either through absorption or contraction, as in jór (q.v.), or through the dissolution or breaking of e, as in jörð (q.v.); again, the i as initial is in most instances caused by absorption; as of n in í (in) and compds; of v or b in íllr (evil) and compds; of d in some compds in í- from ið;—in Gothic there is only a single word (eisarn, i.e. ísarn = iron) with a long í initial.

III. by comparison with other Teutonic languages it is seen that a radical initial i or j has in the Scandinavian been dropped in a few words, while it has been kept in Gothic, Saxon, and German, thus Icel. ár, Goth. jêr, Engl. year, Germ. jahr; Icel. ungr, Goth. juggs, Engl. young; Icel. ok, Goth. juk, Engl. yoke, Germ. joch, Lat. jugum; Icel. ami, ömurligr, and O.H.G. jamar, Germ. jammer; Icel. upp, Goth. jup, Engl. up; Icel. ér (ye), Goth. jus; Icel. ostr (a cheese), cp. Engl. yeast: in two words, jarteign and jurt, both of them probably foreign, the j stands for w: on the other hand, because of the resolution or breaking of vowels (Gramm. p. xxix, bottom), words which in Engl. and Germ. begin with e are in Icel. often to be found under j, thus Icel. jörð (old Scot. yerth) = Engl. earth, Germ. erde: there are also a few stray words,—jata (a manger) for eta, jeta for eta, jeg for eg (ek).

IV. the Icel. í answers to Ulf. ei (rísa, Goth. reisjan), to mod. Germ. ei in zeit, Engl. i as in time, Icel. tími; in early German the diphthongs ei and í were, as in Icelandic, distinguished (zît, îsarn, = mod. zeit, eisen).

V. in mod. Dan. in a few words the Icel. short i is represented by an e, thus Icel. við, liðr, viðr, siðr, biðja, limr, vinr, sin, = Dan. ved, led, ved, sed, bede, lem, ven, sene, probably owing to the fact that the old Danish pronunciation of i was not the same as the present Icelandic.


J is really the tenth letter of the alphabet, but since it is usually regarded as another form of I, K is commonly reckoned as the tenth letter.


K (ká) is the tenth letter of the alphabet; in the common Runes it was represented by (kaun); the Anglo-Saxon k was called ceân or cên = Germ. kien, a pine or fir-tree; but as this was not a Norse word, the Scandinavians represented it by the Norse word nearest in sound to it, kaun (a boil or scab), which bears witness of the Anglo-Saxon origin of the old Norse Runic poem.

B. Pronunciation.—The k is sounded hard or aspirate, the pronunciation varying as that of g does, see p. 186; it is hard in kaldr, koma, kunna, aspirate in kel, kem, kenna, kið, kyssa, kæti, keyri, vekja, etc.; the only difference is that k has the same sound, whether initial or medial, kaka, kíkir, just as in English: in modern Danish the medial k has been softened into g, e.g. Icel. sök, vaka, líka, Engl. sake, wake, like, are in Danish sounded sag, vaage, lige, whereas Sweden and Norway as well as Iceland have kept the old pronunciation.

2. the letter k before t and s is sounded as g, thus okt and ogt, þykkt and þygt, slíks and slígs are sounded alike; and so k is now and then misplaced in MSS., e.g. lakt = lagt, heilakt = heilagt. The spelling and other points referring to k have already been treated under C, p. 93; for qu = kv see Gramm. p. xxxvi. (II. 1. δ).

C. Changes.—The change of initial kn into hn has been mentioned in the introduction to letter H (B. II. 2. γ), where however ‘hnefi’ ought to be struck out of the list: for the changes of nk into kk see the introduction to letter N.

II. according to Grimm’s law, the Teut. k answers to the Gr. and Lat. g; thus Lat. genus, genu, gent-is, Gr. γένος = Icel. kyn, kné, kind, etc.: but in borrowed words no change has taken place, as in Keisari, kista, kerti, kjallari, = Lat. Caesar, cista, cera, cella; the words borrowed in that way are very numerous in this letter, but there are some slang or vulgar words, which seem not borrowed, and yet no change has taken place.


L (ell) is the eleventh letter of the alphabet, and the first of the liquids. In the Runic alphabet on the Golden horn, as well as in the later Runes, it was represented by , called lögr, q.v. (lögr er það er fellr ór fjalli, Runic poem; A.S. lagu), and was, as the form shews, evidently drawn from the Greek or Latin alphabet. In old MSS. a digraph Digraph for ll is often used for ll, see Bs. i. 333 sqq.

B. The l is in Icel. sounded as in other Teut. languages; but ll, after a vowel and not combined with another consonant, had a peculiar sound, almost dlh, thus, gull, fall, hella, kalla, = gudlh, fadlh, hedlha, kadlha. This pronunciation is still observed in Icel. as well as in some provincial dialects of western Norway, Vorse-vangen, Sogn, Hardanger; in some other parts of Norway it is sounded as dd. There are no means of ascertaining with certainty whether the ancients sounded ll exactly as the Icel. at present do, or whether it was not more aspirate than dental (as llh).

2. the peculiar aspirate sound of l before a radical dental is mentioned Gramm. p. xxxvi. (II): thus holt, allt, gult, íllt, hallt, etc. were sounded (and are still sounded) as holht, alht, gulht, ílht, halht; as also in old writers before d, hold, kald, = holhd, kalhd, although in mod. pronunciation the aspirate sound is less perceived before a media than before a tenuis.

C. In some Icel. words the ll is due to assimilation, and answers to Goth. lþ, Saxon and Germ. ld, e.g. Icel. gull = Goth. gulþ, Engl. and Germ. gold; it is however likely that originally these words were distinct in sound from those which had a radical ll, and it may be that the present peculiar sound of ll was due to this cause—that the sound of the assimilated ll prevailed and became universal, whilst the original radical ll sound was lost; though even in the earliest rhymes no distinction is to be perceived.

2. in much later times ðl assimilated into ll in a few words, brálla = bráðla; as also lr into ll in inflexions, hóll = hólr, stell = stelr, Gramm. p. xvi. (I. 3. α): in still later times rl changed into ll, jarl, karl, varla, etc., which in mod. pronunciation is sounded as jall, kall, valla, etc.; but this is not observed in writing, although it is so in early print, as also in MSS. of the 15th century.

All words having a radical initial h (hl) are to be sought for under H; see the introduction to that letter.


M (emm), the twelfth letter of the alphabet, was in the oldest Runic inscriptions (the Golden horn, the stone in Tune) figured , and in the later common Runes Common rune form for M, curved and Common rune form for M, straight, whence later, the top being left open, and ; all these forms being clearly derived from the old . Its ancient name was maðr (a man)—‘maðr er moldar-auki,’ in the Runic poem; but the likeness of to a man with uplifted hands is merely accidental.

B. Spelling, Changes.—The m is sounded as in English and other Teutonic languages: it is usually single in the words fim = five, um = umm, fram = framm, but erroneously, for it is sounded double, agreeably with the etymology. Changes:

1. peculiar to the Northern languages is the interchange of mn and fn when both letters are radical; thus, namn, nemna, = nafn, nefna; samn, samna, = safn, safna; jamn = jafn; somna, sömn, = sofna, svefn; hemna = hefna; stamn = stafn; stomn = stofn; hramn = hrafn: the oldest Icelandic vellums frequently use the mn, in namn, samna, hramn, jamn, but more rarely in other instances, as omn, Blas. 46; emni, Arna-Magn. 301. 3; somna, MS. 623. 34; somninum, O.H.L. 82, 83; sjómn, Pd. 14; it is still sounded instead of fn before d, as in hefnd, proncd. hemd, Hom. 7, 18: stemdi = stefndi: the fn has prevailed in the Icel., and is used in the Editions, as also in modern usage: on the other hand, the Swedish has throughout adopted the mn; thus, Swed. hamn = Icel. höfn, Engl. haven; Swed. lämna = Icel. lifna, Dan. levne; Swed. ämne = Icel. efni; each language has taken its course without regard to etymology, for in some of the words f is radical, in others m.

2. otherwise m and f seldom interchange, as in the threefold form of the particle of, um, umb; himinn, himneskr, and hifinn, hifneskr; nema (nisi), cp. Goth. niba; hvilmt and hvilft, q.v.; as also Mal-kólfr = Mal-kolmr, þjálfi and þjálmi, skelmir and skelfir.

II. in vellums m is dropped in the 1st pers. plur. of verbs before the pron. vér, vit, thus höfu vér, eigu vit, but in mod. usage eigum vit; hence comes the Norse form mér (plur.), mit (dual), by an anastrophe of the v and substitution of the final m from the preceding verb: in öllu-megin, báðu-megin, etc., see megin. In old vellums the A.S.  is used to mark a double m, thus frai = frammi; in most cases a medial or final m is marked by a stroke above the line. The Rune is often used for the word maðr.


N (enn), the thirteenth letter, is in the old Runes represented on the Golden horn by the character , on the stone in Tune by , and in the later Runes by or , all derived from the Lat.-Gr. N; it was called nauð (need, A.S. neâd), nauð görir neppa kosti, Runic poem. In ancient MSS. the capital N or the A.S.  is used to mark a double n, thus, ma, kea, = mann, kenna.

B. Pronunciation.—The n is sounded as in other Teut. languages; but nn after a diphthong has a peculiar sound like dnh, thus steinn hreinn = steidnh hreidnh; whereas, after a single short vowel the sound is as usual, hann, mann; this ndh sound does not seem to be ancient, as may be seen from rhymes such as, seinn þykki mér sunnan, Sighvat: a confusion between rn and nn first appears in MSS. of the 15th century; e.g. eirn hreirn, = einn hreinn, and so in early print: before t the n is aspirate, vint = vinht, cp. introduction to letter L.

C. Changes.—The nn before r in olden times was often changed to and sounded as ð, not only in maðr, suðr (= mannr, sunnr), öðrum, aðrir, aðrar (from annarr), miðr (= minnr), in which cases it is still sounded so; but also in saðr, muðr, bruðr, fiðr, meðr, uðr, guðr, kuðr, = sannr … kunnr; tveðr = tvennr, gryðri = grynnri, Bs. i. 342, 349; saðrar = sannrar, Greg. 23, Líkn. 3: it is so used in rhymes by the poets; in all these latter instances the nn has reappeared in mod. usage; cp. Engl. mouth = munnr, but sunna (the sun). May not the change of the participles in -iðr into -inn (Gramm. p. xxiv, col. 2) be due to the same phonetic principle, but in inverted order? The n is elided in jamn-mikit, sounded and spelt jam-mikit; jam-góðr = jamn-góðr:—nn or n for nd, in sunz = sunds, lanz = lands, munnlaug = mundlaug; bundnir, sounded bunnir and spelt so, Edda i. 240. In some words the nn is due to assimilation, as that of zn in rann, Goth. razna; but often of or nd in the cognate Teut. languages, thus Icel. nenna, Goth. nanþjan; finna, Engl. find. For the absorption of final and medial n see Gramm. p. xxx, col. 1.

For words with a radical h (hn) see under H.


O, the fourteenth letter, is in the oldest Runes, on the stone in Tune, and on the Golden horn figured by , which was evidently taken from the Greek Ω; the later common Runic alphabet in earlier monuments has no ó, but uses u or au instead, e.g. on the Jellinge stone in Denmark. Afterwards the Rune , , or , , appears under the name of óss in the Runic poems—óss er flestra ferða (= fjarða) = all firths have an óss (mouth). The form was evidently taken from the A.S. Runic , which stands for a, and in A.S. is called ós, which answers, not to Norse óss (ostium), but to áss (= ans, i.e. the heathen gods); but the Norsemen or Danes in borrowing the Rune seem to have misinterpreted its name or mistranslated it from ignorance of the phonetic laws existing between the A.S. and the Norse. The in Scandinavian Runic inscriptions is therefore a mark of later date (11th or 12th century).

B. Pronunciation.—The o is either short (o) or long (ó); the former (o) is sounded like Engl. o in cod, the latter (ó) as in Engl. no, note; but the rules given at the beginning of the introduction to letter A (p. 1) apply equally to this letter, bōð being sounded bawth, but krŏss in North. E. cross.

C. Changes.—In most of the oldest vellums o instead of u is used throughout in inflexions, -o, -or, -om, -on, -oð, -ot, -osk, -oll, -onn, instead of -u, -ur, -um … -unn (Gramm. p. xxxv, col. 1, A); afterwards both forms are used indiscriminately, till in the 15th century the u prevailed, and has kept its place ever since; whether there was a difference in sound, and what, we are unable to state.

2. so also in a few root words, goð, goll, fogl, oxi, skolu, monu, hogr, togr, monr (Dan. mon), smogoll, = guð, gull, fugl, uxi, skulu, munu, hugr, tugr, munr, smugull; on the other hand, is sonr (a son), but sunr the older form.

3. a and o or u interchange in the inflexions, fagnaðr, fognoðr, fögnuðr; kallan, kollon, köllun.

4. ú has changed into ó in the prefixed negative, ó-vitr for ú-vitr (unwise).

5. into jó, njóta, originally njúta; ljós, Swed. ljús: forms like mjókr for mjúkr, dókr for dúkr may also be found in vellums, but are very rare.

II. ́, the vowel-change of á (see p. 1), is frequently spelt o (tor, nott, = t́r, n́tt), but was in sound different from ó proper, and has since disappeared from the language, although remains of this ‘umlaut’ still exist in nótt, ól, spónn (= n́tt, ́l, sp́nn), but this o is sounded exactly like common o. So also o and ö are confounded in MSS., bornom = börnom = börnum. For the absorption of consonants see Gramm. p. xxx, col. 1, and the words themselves.

Owing to the inability of the Scandinavian languages to sound v (w) before a vowel of the u class, several root words, which in dictionaries of the cognate languages (Germ., Engl.) begin with w, are in the Icelandic to be found under o, as okr, orð, orka, ormr, Óðinn, óðr, ósk, ómr; as also j, in ok (jugum), ostr, and ok the conjunction.


P (pé), the fifteenth letter, was not figured in the old Runic alphabet, in which the bjarkan () was made to serve for both b and p; it is found only in very late Runes, as e.g. the Runic alphabet of the Danish king Waldemar (died a.d. 1241), where it is figured or as a dotted (ᛔ), Skálda 177, and the Arna-Magn. Runic MS.

II. the p is in Icel. sounded as in Engl., pína = pain, hlaupa = leap.

B. Remarks.—As all words with p initial have been borrowed at different times from foreign languages, the number of them goes on decreasing the farther we go back into antiquity; this is also the case in other Teutonic languages; the vocabulary in Ulf. presents about seven p words,—paida, plapja, plats, plinsjan, pungs, prangan, pund; the old A.S. poems about the same number,—plega, plegjan, pæd, pund, pynd, pyt, pad, peord (while the oldest and best, Beowulf, has none), see Grein. The ancient Icelandic or Norse poems of the heathen age have still fewer than the A.S.; the first words we meet with are penningr, a penny, Bragi; pundari, Egil (see ljóð-pundari);—which, with some other words beginning with p, are from the heathen age. Along with the introduction of Christianity many such words came in, chiefly through the English, e.g. prestr, pína, pínsl, páskar, páfi, pistill, prédika: through trade from the Norman-English, prúðr, prýði, páll, pallr, pell, poki, partr: and lastly, through the English trade with Iceland in the 15th and 16th centuries, prenta, púðr, petti, peisa, etc.: some few words, too, have since been adopted from the mod. Danish. A few words may be traced to Gaelic, and a few have been traced to the Chudic (Finnish); the scantiness of such words, however, shews better than anything else how very small indeed was the influence these languages had on the Norse, all the more so as the Finnish vocabulary abounds with p words. The letter p in an Icelandic Dictionary stands quite apart from all the other letters, for it is made up of a motley collection of words, incoherent and broken, containing no roots, no great verbs, particles, or such words as make the stock of a genuine vocabulary. The absence of initial p in the Teutonic language is not due to any inability to pronounce it, but to causes inherent to the parent language of the Teutonic as well as the classical languages, for in Greek and Latin the letter b, which, according to Grimm’s law, answers to the Teutonic p, stands exactly in the same predicament as p in the Teutonic; there is no single instance of ‘lautverschiebung’ from a Gr.-Lat. b to a Teut. p (Curtius): no word beginning with p is formed by ‘ablaut,’ and only a few are derived by ‘umlaut’ (prýði, pyngja, pæla). For other details see the introduction to letters B and F.


Q (kú), the sixteenth letter, was in old vellums chiefly or only used in the digraph qu, perhaps under the influence of the Latin; it was afterwards disused in MSS. of the 15th century, but was reintroduced in printing (even Björn Halldórsson’s Dict. has a special Qu), until of late it has been discarded, and k is used throughout. All words beginning with Qu are therefore to be sought for under k.


R, (err), the seventeenth letter, had in the old Runes two forms; one as initial and medial (radical), ; the other as final (inflexive), or . Of the last two, is used in the old Runes (stone in Tune, the Golden horn) in the words gastir, hokingar, wiwar, as these inscriptions have now finally been read and settled by Prof. Bugge of Christiana; is used in the common Runes; and its name was reið,—reið kveða rossum versta, in the Runic poem.

B. Pronunciation, Spelling.—The pronunciation is as in Italian or in mod. Gr. (rh), and this still survives in Norway and Sweden, whereas the Danes have adopted a guttural r, which an Icelandic throat is unable to produce. In ancient times radical and inflexive r were perhaps different in sound, as may be inferred from the spelling on the old Runic monuments, as well as from comparison; for the inflexive r was in the Gothic a sibilant (s), so that the Runic ᛉ and ᛦ may well have represented a sound intermediate between r and s.

II. the inflexive r is assimilated in words such as heill, steinn, lauss: dropped in nagl, fors, son, vin, see the Gramm. (xvi I 3 α; xix I and I 4; xx, degrees of comparison II):—the ancient writers have a double r in nouns and adverbs, such as sárr, stórr, ferr, síðarr, optarr, meirr; even against etymology, as in hárr (high), márr (a mew). In mod. usage a final rr is never sounded. Again, in gen. and dat. fem. and gen. plur. and in compar., in words such as þeirri, þeirrar, þeirra, færri, fárra, the mod. sound and spelling is rr, where the ancients seem to have sounded one r only, þeiri, þeirar, þeira, færi, fára, which pronunciation is said to be retained in eastern Icel.; the Editions, however, have mostly adopted rr. The spelling of the vellums is often dubious, as in them a double r is written either dotted () or with a small capital ʀ, but mostly without a fixed rule:—Norse vellums often give rs for ss (mersa = messa, þersi = þessi, e.g. in the Hauksbók).

C. Changes.—As the Icel. cannot sound w before r, a set of words which in Engl. and even mod. Dan. and Swed. begin with w, in Icel. belong to r; thus, rangr, röng, rata, reini, reitr, reista, reka, ríða, ríta, reiðr, rindill, risi, rist, röskr, róg, rugl, rölta, qs. wrangr, … wrölta. In a few words the r has been dropped after a labial, thus Icel. víxl = A.S. wrixl, Icel. beisl = A.S. bridels, Lat. frenum, Icel. bauta-steinn qs. brautar-steinn; Icel. vá qs. vrá, Hm 25, Skv. 3. 29; or a false r is inserted, as in the Icel. ábristir = Engl. beestings, Goth. beist. Germ. beist-milch; bræla and bæla, bál; analogous are Engl. pin and prin, speak and A.S. spræcan, Germ. sprechen, Icel. freta, Lat. pedo:—in a few Norse vellums ðr for ð is used before s, l, n, oðrla = öðla = óðala, öðrlask = öðlask, Guðrs = Guðs, heiðrnir = heiðnir, liðrsemd = liðsemd, soðrla = söðla, ráðrleitni = ráðleitni, e.g. the O.H.L. (see the pref. to Prof. Unger’s Ed. p. ix), owing to an inability of sounding ðl, ðs. Again, metathesis has taken place in ragr, rass, = argr, ars.

All words having a radical initial h (hr) are to be sought for under h; see the introduction to that letter.


S (ess), the eighteenth letter, was, in the old Runes, on the stone in Tune, and the Golden horn, figured ; in the common Runes ; in the latest Runic inscriptions (12th and following centuries) or . Its name was ‘Sól’ (Sun)—Sól er landa ljómi, in the Runic poem; the was specially, from its form, called the ‘kné-sól’ (knee-sun).

B. Pronunciation, Changes.—Sounded sharper than in English. The s is in mod. Icel. pronunciation the only sibilant sound; in olden times s and z were distinguished in sound as well as in writing, but afterwards the z sound was lost or assimilated with s.

II. Changes: s into r, as vera, var, er, for vesa, vas, es; as also the particle es for er; Gothic s into Scandinavian r in the words, Goth. hausjan, auso, = Icel. heyra, eyra; the inflex. Goth. -s into Scandin. -r: an assimilation has taken place in such words as laus-s, ís-s, for laus-r, ís-r: again, in vellums, ss for s in such forms as búss (gen.) from bú, nýss = nýs, hirðiss = hirðis (gen.): in mod. usage this inflexive s is dropped in sound and spelling, laus, ís: the ancients, on the other hand, said víssa, víssi, mod. vísra, vísri (sapientum, sapienti):—sn is sounded stn, stnúa, stnöri, stnöggr, stnjór …, = snúa, snöri, snöggr, snjór …, and thus spelt in some Norse vellums (e.g. the Barl.): here come in also such forms as laustn, njostn, ristna, = lausn, njósn, risna, reistn and reisn, O.H.L. (pref.) ix; so also the forms Ást-ríðr, Ást-leifr, Ást-lákr (see the remarks s.v. ást), = Ás-ríðr … Ás-lákr, Baut.

2. skl = sl, thus sklakka = slakka, D.I. i. 280, l. 10, but rare: cp. the Germ. spelling schl = Icel. and Engl. sl (Germ. schlagen = Icel. slá); as also the Fr. esclave and slave.

3. sk corrupted into skr, skokkr, skykkjum, and skrokkr, skrykkjótt; analogous are Icel. skjallr, Engl. shrill:sk for s, in sjaldan and skjaldan, Icel. saur-lífi, Dan. skör-levned.

4. sk answers to Engl. sh (skip, fiskr, = Engl. ship, fish), except in a few words, as Engl. skin, score, which may be borrowed from the Norse.

There are more words beginning with s than with any other letter of the alphabet; this is due to the combination of sk, sm, sp, and st.


T (té), the nineteenth letter, was in the Runic alphabets represented by , and in later Runes also by ; its name was Týr—‘Týr er einhendr Ása,’ in the Runic poem; a marked , ‘stunginn Týr’ represented the d. Týr was the first of the third and last group in the alphabet, T b l m y, which was therefore called Týs-ætt, or the family of Tý, cp. the introduction to letters F and H.

B. Changes.—T is sounded as in English. Various kinds of assimilation take place with this letter,—dt into tt, e.g. neuters of adjectives, gótt, ótt, blítt, qs. góð-t, óð-t, blíð-t, see góðr, óðr, blíðr: ht or kt into tt, sótt from sjúkr (Goth. sauhts), see Gramm. p. xxx, col. 1: nt into tt, stuttr, brattr, vöttr, for stunt, brant, vant; vittr (i.e. vetr) for vintr: ndt into tt, as statt, batt, bitt, vatt, hritt, hratt, qs. standt, bandt, bindt, from standa, binda, vinda, hrinda: the mod. preterites, benti, lenti, synti, kynti, from benda, lenda, synda, kynda, qs. bend-ti, lend-li, where the ancients have bendi, lendi, etc.: into t or tt, as imperatives, viltú, sittú, vittú, vertú, for vil-þú, sit-þú, vit-þú, ver-þú: also in mod. pronunciation, tótt for topt, bátt for bágt: tt for t after a long vowel or diphthong, mjó-tt, fá-tt, há-tt, smá-tt, ný-tt, from mjó-r, fá-r, há-r, smá-r, ný-r: kð, pð, into kt, pt, sekt, vakti, dýpt, dreypti; older and better, sekð, dýpð, see introduction to letter D. In some Norwegian vellums a digraph is used for ð, etða, matðr, atðrum, þatðan, smitðja, ytðru, = eða, maðr, aðrum, þaðan, smiðja, yðru, see Þiðr. (pref. xvi): also stn for sn, stnúa, stnjór, for snúa, snjór, see introduction to letter S.

II. an initial t, as is remarked by Prof. Bugge, has become k in the words kvistr, kvísl, qs. tvistr, tvísl (from tví-).

III. following Grimm’s law the Teut. t answers to Gr. and Lat. d, δυσ- = tor-, δάκρυ = tár, δόρυ = tré, δέκα = tigr, Lat. dŏmo = temja, videre = vita, sedere = sitja, and so on.

2. the Norse t, as well as Engl., answers to High Germ. z, Icel. tíð, tal, = Germ. zeit, zahl, etc.


U (ú), the twentieth letter, was represented in the Runic alphabet, both on the stone in Tune and in the later Runes, by , and was called úr, Skálda 176,—úr er af eldu járni, the Runic poem: u is sounded like eu in Fr. feu, ö in Germ. hören; ú like oo in Engl. root. In mod. Engl. the Icel. ú is represented by ou, ow, e.g. Icel. út, hús, búr, = Engl. out, bouse, bower, such words being in Early Engl. written ût, hûs, etc.; they still retain their Scandinavian pronunciation in North England. For the changes between o and u and ó and ú see the introduction to the letter O, p. 462. As with o, so with u, the words with initial v have, in the Scandinavian languages, dropped that letter, e.g. una = wone; undr = wonder; und = wound; ull = wool; úlfr = wolf.


V (vaff), the twenty-second letter of the alphabet, was by the ancients called vend, q.v. = A.S. wen, whence Icel. vindandi, q.v.; like u, it is represented by in the Runes; in old vellums by , a form borrowed from the A.S.; later by v or u, according to the Latin usage. The Icel. v answers etymologically to Engl. and Germ. w (not v), but the form w is little used, though not quite unknown in Icel. vellums.

B. Pronunciation, Changes.—The v is now sounded in Icel. as in English, by a slight touch of the lip and teeth (not like the South German, with both lips). It is said (Rietz) that a w or double v is still sounded in a remote Swedish county (Dalarne), and the opinion of English phonetic philologists is that the English w represents the old Teutonic value of that letter, which has since been lost all over the Continent, as well as in Iceland. The Icelandic formation of words goes far to prove that the old Scandinavian v was a semi-vowel, and not a full consonant as it is now:

1. traces are found of v alliterating with a vowel; on an old Runic stone (of the 10th century?) in the island Öland (off the coast of Schonen) we read—Vandils jörmungrundar, ur-grandari (v, i, and u making the three staves). In verses which have passed through Icelandic oral tradition, alliteration like this could hardly have survived, except in a very few cases (there are, however, some such, svaf vætr Freyja átta nóttum, Þkv. 28; Óðinn á jarla þá er í val falla, Hbl. 24; ónu verr, Ls. 36, cp. Hm. 22); but on the Runic stone, the words still remain as they were first engraved.

2. the frequent ‘vocalisations’ involving the loss of v, which is indeed the most mutable of all letters:

α. ve, vi change into y, vá into ó: a v cannot be sounded before a u-vowel, viz. before o, ó, u, ú, y, ý, œ, ø; countless instances of this are to be found under the heads of v-, dv-, hv-, kv-, sv-, tv-, þv-; cp. as specimens the tenses of verbs, vaða, valda, vaxa, vefa, sofa, koma, vinna, vinda, svimma, svella, vella, velta, verpa, verða, hverfa, svelta, svella, sverfa, þverra, svelgja, þvá, sverja; and also hváll and hóll, hvat and hót, ván and ón, váru and óru, kvef and kóf kaf kœfa, svartr and sorti surtr syrta, verk and yrkja orka.

β. more rarely, before other vowels, as, þeita for þveita, silungr for svilungr, hika for hvika, skak for skvak, þi and því.

γ. v and j interchange, as in hjól for hvel, sjót for sveit, skjal for skval, jurt for vurt, jartegn for vartegn, hvern for hjarni; in verbal inflexions, -va into -ja, as byggva into byggja, syngva into syngja.

δ. in a few words the v has been saved by a change in the following vowel, as in verk, cp. A.S. weorc, virkr for vyrkr; virgull (a halter, Goth. wurgils), væðr and œðr, Svænskr Sœnskr.

ε. for the loss of v before a u-vowel see the introduction to the letters Ó, U, Y, Æ, Ö; so in parts of England and in Scotland at the present day men say ool for wool.

ζ. in a few other words initial v is dropped when in compds, vegr in Nor-egr, einn-ig, hinn-ig, þann-ig, hvern-ig; vangr in kaup-angr, Harð-angr; völlr in þreskj-öldr; valdr in Arn-aldr, Har-aldr; verðr in dög-urðr; vindr in Ön-undr; vin in Björg-yn; vé or veig in Þyri, qs. Þór-vé, and in Dan. Odens-e.

η. again u has changed into v in várr qs. úrr, órr, A.S. ûre, Engl. our, and in vesall qs. usall (see várr and vesall).

II. changes of later date, in the 14th and 15th centuries, or somewhat earlier,—old Icel. was turned into vó, and at last into vo; svá, svó, svo; vár, vór, vor; tvá, tvó, tvo; ván, vón, von; vápn, vópn, vopn; vátr, vótr, votr; váði, vóði, voði; kváma, kvóma, koma: the old is proved by rhymes, as vátr and gráta, svá and á; that this did not change immediately into the present vo, but passed through an intermediate vo, is shewn by rhymes in poems of the 14th and 15th centuries; e.g. rr and Þórr, Skíða R. 47, 70, 119, 181; stór, vór, 69; vóðinn, Óðinn, 109; vótt (testem), vórt, 122; góma, kvóma, Völs. R. 199; vórr, stórr, 212; stórr, hvórr, 248: the still remains in vóru (erant), although short o in voru is rapidly displacing the old long vowel. This later change of into vó, compared with the old dropping of the v whenever it came in contact with a u or o, shews that at the time when it took place (the 14th century), v cannot any longer have had the same sound as it had five or six centuries before, when the great and systematic vocalisation of it took place. In mod. Icel. v has even reappeared in a few, especially verbal, forms (where people are still conscious of the lost v), so that Icel. now say hvurfu, hvorfinn, hvyrfi (from hverfa); so also, but esp. in later vellums, less in speech, the forms vurðu, vorðinn, vultu, vundu, vorpið, etc., from verða, velta, vinda, verpa; cp. also the mod. sound of the word Guð: again, in words like ull, úlfr, orð, una, etc., a restoration was impossible, all remembrance of the v having been lost for a thousand years: but phonetically, since v became a labial consonant, an Icelander might and could say vull, vúlfr, vorka, vorð, etc., just as well as von, vor, votr.

III. for the dropping of v before r (and l) see the introduction to the letter R (as in rangr = Engl. wrong): it is doubtful whether Icel. vág-rek (= a wreck, flotsom) bears any relation to vágr (a wave): the v may here have been saved by means of a false etymology, vagrek for vrak.

2. in a few cases an aspirate (h) has been substituted for an original w, e.g, Icel. hreiðr (a nest), cp. Engl. wreath, Goth. wriþus; Icel. hrista, akin to Engl. wrist. Dan. vriste; Icel. hrekja, akin to A.S. wrecan, Engl. wreck; Icel. hrína, akin to Dan. wrinske; and perhaps a few more words.

3. in still fewer instances the r has fallen out, the w or v remaining; these words are veita (to trench), veiting (a trench, drainage), for vreita, vreiting (akin to wríta); veina (II) = vreina; and lastly, vá for vrá (a cabin).

IV. an interchange of v and f occurs in a few instances, e.g. ái-fangi, áfangi, qs. ái-vangr; in var-nagli and far-nagli; in varinn and farinn, see fara A. VI. β; in válgr and fjálgr.

2. in inflexive syllables, like örfar, snjófar, bölfi (ör, snjór, böl), and the like, the change of v into f is etymologically erroneous, but phonetically indifferent, final or medial f being one in sound with v.

V. for the v or u as the cause of a vowel change, see Gramm. p. xxix.

2. it is dropped in inflexions in many words, such as in mörr, böð, stöð, dögg, högg, böl, öl, söl, fjör, smjör, mjöl, kjöt, hey, sær, snær, fræ, bygg, lyng; adj. hár, mjór, þjökkr, dökkr, röskr, glöggr, etc.

Many of the preceding phenomena (esp. in I. and III) could not possibly be accounted for, unless we assumed that, at some early time, when those changes took place, the v was sounded, not as a consonant, but as a kind of oo sound, half consonant, half vowel; if so, no sound could answer more nearly to it than the mod. Engl. w; the change may have taken place at a very early date, prob. before the settlement of Icel. Norse words in the Shetland and Orkney dialects point to v not w, e.g. voe = Icel. vágr.


X. (ex) commonly represents ks, gs (as in other languages), where both letters are radical, thus, ax, fax, lax, sax, öx, vax, vaxa, sex, uxi, vöxtr, fox, jaxl, öxl, q.v.; but hugsa from hugr; lags from lag; loks from lok; oks (gen.) from ok (jugum); rakstr, bakstr, from raka, baka, etc. The vellums use x in other cases, e.g. sterxti = sterksti, the strongest, Clem. 146; tax (gen.) from tak, N.G.L. i. 47; dúx = dúks, Clem. 127, l. 8; lox = loks, 134; vitrleix = vitrleiks, 142; almattex = almáttigs, 133; víxla = vígsla, N.G.L. i. 9; fulltinx = fulltings, ÓH. 242; vaxcliga = vaskliga, Mork. 178; lyxc = lýksk, Íb. (fine); fexk = fékksk (from fá the verb), Bs. i. 351; ux = ups (q.v.), N.G.L. i. 368: or again, vegs = vex (the verb), Hm. 119; lags = lax (salmon), Sæm. 212, l. 20 (Bugge); dax = dags, N.G.L. i, 23; but on the whole the vellums distinguish gs, ks, and x, shewing the pronunciation in olden times to have been more distinct than it is now, when all three forms (gs, ks, x) represent the same sound, no matter whether the s be inflexive or not; thus in common modern spelling, both hugsa and huxa, dags and dax are used at random. In vellums x and r are very much alike: hence in the well-known passage in Vsp. the misreading of sarum (sordibus) for saxum (ensibus), in all Editions, until Prof. Bugge noticed the stroke underneath the line in Cod. Reg.


Y is of later origin, and only found in derived words, being an ‘umlaut’ from u, (y from u, ý from ú); in the Runic alphabet it is placed at the end, and marked , see Skálda (ii. 72); it is there called ýr, a yew-tree,—ýr er vetrgrænst viða, ‘ýr’ is the ‘winter-greenest’ of trees, Runic poem.

B. The independent sound of y, ý is now lost in Icel., being replaced respectively by the sounds i, í, whereas in mod. Dan., Swed. and Norse the old sound has been preserved; the old Icel. MSS., as well as the rhymes in old poems, distinguish both, except in a few instances, see Gramm. p. xxxv, col. 2 (η). The change from y to i seems to have begun about the time of the Reformation, but in the first printed books, e.g. the N.T. of 1540 and the Bible of 1584, the distinction is still well kept, the remembrance of the old form and etymology being then still alive. Later, the writing became very confused. Some transcribers of the 17th century, e.g. Ketil Jörundsson, a noted copier of old vellums, took the better course, never writing y at all, but i throughout; the same may be observed in the handwriting of some Icelanders down to the present day. In printed books of the 17th and 18th centuries the confusion is great, till of late an accurate spelling has been re-established, though even this fails in a few words; e.g. the ancients spell þrysvar, gymbr, q.v.; the mod. þrisvar, gimbr. The poets of the last three centuries make i and y, ei and ey rhyme indifferently, according to the usage of the living tongue.

II. an initial v is dropped before y, as in yndi, yrði, yrkja, etc.


Z (zet). The ancient language had two sibilant sounds, s and z; of which the z never stands at the beginning of a word, but is merely an s assimilated to a preceding dental, in the combinations ld, nd, nn, ll, rð, gð, see Gramm. p. xxxvi, col. I. β: its use in ancient vellums is very extensive:

1. in genitives; trollz, íllz (íllr), allz (allr), holtz, Skm. 32; gullz, 22; ellz = elds, botz = botns, Gkv. 3. 9; vatz and vaz = vatns; keyptz, Hm. 107; mótz, Knútz or Knúz = Knúts; vitz (vit); orðz, sverðz, barðz, borðz, garðz, harðz, langbarz, Gkv. 2. 19; Hjörvarðz, Hkv. Hjörv. 19; morðz, bragðz, flagðz, Frissb. 107, l. 19; or also orz, Hm. 141, etc.; prestz, Christz, passim; tjallz, Edda ii. 314; landz or lanz, passim; fjallz, Edda ii. 339; but tjalldz, 527; elldz, vindz, 317, 318; gandz, 525; brandz, 529; valldz, 338; sverðz, borðz, 331; but borz, 462, 1. 20; garz, 529; loptz, 341 (twice); but lopz, 317; netz, 327; gautz, 345; hugskozins, Post. 251.

2. in special forms; stendz, Grág. i. 501 (from standa); stennz, id., Ó.H. 143; bitzt from binda, Post. (Unger) 154; vizk, vizt, vatzk from vinda (II), q.v.; but vinnz from vinna, q.v.; biz = biðsk from biðja, Post. (Unger) 240: indeed bizt, bazt may be both from binda and biðja: bletza and blezza (to bless), höllzti, q.v.; beztr or baztr, the best; œztr = œðstr; þatz and þaz—þat es, Sæm. passim; þatztu, Am. 87; hvártz = hvárt es, Grág. (Kb.) i. 161: even mz (or mzt) for the older mk, þóttumz, Gkv. 2. 37.

3. when the z is due to a t following it; in the reflex, -sk is the oldest form, whence -zt, -z, -zst; andask, audazt, andaz, andazst: in the superl. zt, efztir, Frissb. 78, 1. 20; harðazta, l. 33; snarpazta, l. 16; ríkaztr, 207, l. 18; fríðuzt, l. 34; hagazt, Vkv. 18; grimmaztan, Edda ii. 530; máttkaztr, 280; hvítaz, 267; but st is the usual form, thus, sárastr, grimmastr, hvassastr, Gh. 17: in Ázt-ríðr = Ást-ríðr, Ó.H. 198, l. 12.

4. in such words as veizla, gæzla, reizla, leizla, hræzla, gæzka, lýzka, æzka, æzli, vitzka or vizka, hirzla, varzla, hanzki, = veitsla, … hirðsla, varðsla, handski, etc.: in reflex, neut. part., thus, hafa borizt, komizt, farizt, tekizt, fundizt, glazt, sagzt, spurzt, kallazt, dæmzt, átzt, … (from bera … eiga): in reflex. 2nd pers. pl. pres. and pret., e.g. þér segizt, þér sögðuzt, qs. segit-st, sögðut-st, so as to distinguish it from the 3rd pers., þeir sögðust, qs. sögðu-st.

5. Gitzurr or Gizurr, Þjazi, Özurr; afraz-kollr, Ó.H. (pref.); huliz-hjálmr; Vitaz-gjafi, q.v.; but alaðs-festr, Grág. (Kb.) i. 88; viz, see víðr II: in foreign names, Jariz-leifr, Jariz-karr, Buriz-leifr, Gkv. 2. 19, Fms. vi. The etymology of words may often be decided by this; e.g. in beisl, a bridle, beiskr, bitter, the s of the vellums shews that neither word is derived from bíta; beiskr is in fact akin to Engl. beestings, Ulf. beist = ζύμη, A.S. beost: geiska fullr, Hkv. 2. 35, is not from geit, but from geisa: laz or latz (p. 376, col. 1) is from Fr. lace, not = Icel. láss: misseri (q.v.) is no relation to miðr, etc.: at lesti, at last, being spelt with s, not z, is not related to latr, but derived from leistr = a cobbler’s last, at lesti = Lat. in calce, see Mr. Sweet’s Ed. of Gregory’s Pastoral Care, p. 474: again, vaztir is akin to vatr = vatn: exceptional cases,—vissi, pret; from vita, and sess, a seat.

II. after a single dental (unless it be t) s, not z, is written; thus, gen. Guðs, boðs, brauðs, auðs, góðs, óðs, vaðs, liðs, öls, fals, háls, frjáls, víns, eins, etc., passim: z is quite exceptional, e.g. liðz, Frissb. 106, ll. 16, 33 (but liðs, Hbl. 33, Am. 43): so also after rn, rl, nl, rn, fn, gn, barns, Clem. 134; karls, Hkv. 2. 2; jarls, Hm. 97; hrafns, segls, regns, tungls (regns, Edda ii. 340). The vellums are very irregular in the distinction of a single or double consonant, but the sibilant used shews the true form of the word; in ‘Odz Colssonar,’ Ó.H. (pref.) l. II, the z and s shew the names to be Oddr and Kolr, not Oðr, Kollr; in a vellum els would be gen. of él, ellz of eldr; in grunz, Edda ii. 287; lunz, 317; hlunz, ranz, lanz, 333; elz, Post, (Unger) 234; golz, 225, l. 23; odz, Ó.H. (pref.), l. II; alz, etc., the z shews that though there is only one n, l, etc. written, they were actually sounded double, grunnz, hlunnz, rannz, landz, eldz, gollz, oddz, allz.

2. the s does not change into z if the word is a compd; as, skáld-skapr, vind-svalr, út-suðr, passim; hirð-stjóri, Edda ii. 335, shewing that in ancient times the pronunciation was more distinct than at the present day; the z in orðztír (Edda ii. 344, orztír, 463) shews that the word is qs. orðz-tírr; yet we find such forms as innzigli, Post. 238; guðzspjall, 239; ástzamliga, 243; handzceld, Barl.; randzaka, Post. 134, l. 29; but rannsaka, l. 14; nauzyn = nauðsyn, Skálda 167. 21; nauzun, Edda ii. 236; anzvara, annzkoti, = andsvara, andskoti, etc.

III. about the 15th century (or earlier) the z sound began to disappear, and s took its place, being at present the only sibilant used in Icel. In later vellums the z is therefore either little used or is misapplied, as in the additions by the third hand in the Flatey-book, or it is used to excess as in modern Dutch. In modern spelling, including Editions of Sagas, the z has been disused, except in the instances coming under the rule given in I. 4: yet with exception of ðs, for the moderns write leiðsla, hræðsla, beiðsla, náðst, old leizla, názt, except in reisla (i.e. reizla) from reiða; hirzla qs. hirdsla.

2. zz is sounded as ss, blessa, Gissur, Össur; so also vass, boss, = vatz, botz; even ors, gars, lans, sans, for orz, garz, lanz, sanz (gen. of orð, garðr, land, sandr).


Þ (þorn) was adopted from the Runic alphabet; its ancient name was þorn (thorn),—þann staf er flestir menn kalla þorn, Skálda (Thorodd) 168, cp. Edda ii. 365,—and it is still so called in Icel.; the ancients also called it ‘þurs’ (giant), which was originally the name of a magical Rune, intended to cause love-madness, and in the Runic poem it is so called—þurs veldr kvenna kvillu; but in the poem Skm. ‘þurs’ means the magical Rune,—‘þurs’ ríst ek þér ok þrjá stafi, ‘ergi,’ ‘æði,’ ok ‘óþola,’ Skm. 36. Thorodd proposed to call it ‘þé’ (like dé, té, bé), Skálda 168. In the Runic inscriptions it is marked , seldom Angled form of rune Thorn; the letter is evidently derived from Gr.-Lat., being a Δ or D with the vertical stroke prolonged both ways.

B. Spelling, Pronunciation, Changes.—For the spelling of the ancient vellums see introduction to letter D (p. 93, col. 2). In Icel. there is phonetically a double th sound, as in English, but subject to a different rule; the hard th, marked þ is only sounded as the initial letter of distinct syllables; whereas the soft th, marked ð is only sounded as a medial or final; and that the case was the same in olden times, as early as the 12th century, is borne out by the statement of the second grammarian (Gramm. p. xv, col. 1), who counts hard th, or þ, among the ‘head-letters,’ as he calls them, whereas the soft ð he counts among the ‘sub-letters’ (p. xv, col. 2, ll. 4–6). That the initial th had only one sound in Icelandic is also borne out by the mod. Faroe dialect, which has the closest affinity to the Icelandic; for here the initial þ has, in pronouns and particles as well as in nouns, changed into t as in ting, tu, teir. But in the rest of Scandinavia the case is different, for there (Dan., Swed., Norse) the initial þ has been changed into d in all particles and pronouns, de, du, der, dem, den, dette, dig, deden, for-di (ti is an exception); whilst, in all other words, it has been changed into t, as in ting taale, tre, etc., which points to a hard and soft th sound, used not as in Icelandic, but as in modern English. According to the views of a gradual and successive ‘laut-verschiebung,’ as set forth in Mr. Sweet’s essay ‘On the Old English Ð’ (Appendix 1. to Gregory’s Pastoral Care, p. 496 sqq.), the Icelandic and the Faroïc represent phonetically a later, the early Danish (old Scandinavian and English) an earlier stage in the development of this sound. It is curious to see how in the Faroïc the sound has come round to Gr.-Lat. again; thus Faroïc trir, tu, = Lat. tres, tu, in Dan. tree, but du.

II. in Icelandic a word with initial þ forming the latter part of a compound, or even if spelt separately, is apt to be changed into ð as soon as it loses its full sound, and is pronounced rapidly as an inflexive syllable, the latter part in questions becoming half enclitic, see introduction to letter D, p. 93, col. 2 (C. II). In vellums this is very frequent in the words al-ðingi, Svi-ðióð, al-ðýða (= alþingi…); so also á ðingi = á þingi, Js. 39; örvar-ðingi, id.; Vaf-ðruðnir, Sæm. (Bugge); hug-ðekkr, Ó.H. 16, etc.; the pr. names Hall-dórr, Hall-dóra point to a Hall-ðórr, Hall-ðóra, = Hall-þórr, Hall-þóra; so also Stein-dórr = Stein-ðórr = Stein-þórr, for a þ could only change into d through ð; in Arnórr, qs. Arn-þórr, the þ has been dropped (Arn-þórr, Arn-ðórr, Arn-órr?); lítt-at = lítt-þat, hítt-ó-heldr = hitt-þó-heldr, flýttier, make haste, already cited in Run. Gramm.; cp. also the change of the pron. þú into -du, -ðu, -tu, -ú, when suffixed. Quite different and much older is the dropping of initial þ (i.e. ð) in the particles enn = ann = þann, Engl. than, and in at = þat, Engl. that, Old Germ. daz: in the pronouns þér, þið, for ér, ið, the þ comes from the termination of the preceding verb. For the rest see the introduction to letter D, to which we may add that a single Icelandic vellum, the later handwriting in Arna-Magn. 645, now published in Post. (Unger) 216–236, is interesting for its uncertain use of þ and ð; at the time it was written, the ð was still a newly adopted letter, and the transcriber uncertain as to its use, so that no conclusion may be drawn from this isolated case; these are the instances,—upp ðu, 216. ll. 19, 27, 219. l. 39; skírþr ðegar, 217. l. 9; upp ðegar, 220. l. 1; blezoþu ðeim, 217. l. 34; af ðeim, 223. l. 10; fyrir ðeim, 224. ll. 14, 18; boþer ðeim, 228. l. 19; viþ ðú, 218. l. 13, 235. l. 5; þá ðaþan, 235. l. 17; af ðvi, 219. l. 15, 232. l. 21, 234. l. 11, 235. l. 13; ifer ðá, 222. l. 31; firir ða trú, 232. l. 34; frá ðér at þú (sic), 226. l. 23; frá þér ef ðu (four lines below); ek biþ ðik, 227. l. 17; viþ ðik, 236. l. 7: after a comma, ðá er rétt, 231. l. 36; ðu laust, 233. l. 32: with nouns and verbs, of ðorp ok borgir, 217. l. 35; ok ðökkuþu, 224. l. 25; firir ðys alþyðo, 227. l. 12.

III. the Icel. þ answers to Gr.-Lat. t, see e.g. the root tan (τείνω, tendo, tenuis) compared to the Icel. þenja, þunnr; þrír = Lat. tres; þrömr = Gr. τέρμα, Lat. terminus; þefr, cp. Lat. tĕpidus, etc., see the special words.

2. again, Germ. d answers to Icel. þ, ding, drei, denken; in a few words the laut-verschiebung is irregular, thus, Engl. tight, Icel. þéttr; þurfa = Engl. dare. Only a few words with initial þ have been adopted in later times, such are, þenkja, þanki, þrykkja (= Germ. denken, ge-danke, drücken); these words were borrowed about the time of the Reformation, probably from German, not Danish, i.e. from words with d; in these words the laut-verschiebung, strange to say, has been duly observed, as if by instinct, which would hardly have been the case had it been borrowed through the Danish t: but in tráss = Germ. dratzen, mod. Germ. trotzen, Icel. þrátta, the true form has not been restored; so also in mod. usage Icelanders are beginning to say tak, tak (= Dan. tak = thanks), unmindful of their own þakk, þakka: t and þ are unsettled in tyrma and þyrma; tolla, see þola (II); tremill and þremill; þeisti and teista: f and þ interchange in Icel. þél, Engl. file; þel and Lat. pilus, þel and fjöl, and in a few other words: s and þ in súst for þust.

Æ (Œ)

This is a double letter—æ, compounded of á and e (a + e), being a kind of appendage to á, and œ, compounded of ó and e (o + e), being an appendage to ó. In the alphabet of Thorodd the former was marked ę́, the latter ́, as in vę́n, f́tr; later, the accent was dropped. The ́ is also written ø, , œ, or even o or eo, thus føti, fti, fœti, or eorri = œrri, whence in modern print æ, œ: most vellums write æ (a + e) and œ (o + e) respectively, which characters have been adopted in modern print. In Icel. the œ-sound was soon lost; only the earliest vellums distinguish the two sounds; then in later vellums ę,  are used indiscriminately, the difference in sound being lost, the characters remaining, till at last the useless sign disappeared. Almost all the Icel. Sagas are preserved in vellums later than the time when the sounds had become confounded; the Cod. Reg. of the Grág. still keeps the distinction, owing probably to its excellent old originals; the Cod. Reg. of the Sæm. Edda uses both signs, but misplaces them, thus, Hm. 92 męta, but fr (pres. from fá), in the same verse. The confusion between æ and œ is purely Icelandic, for in Denmark, Norway, and Sweden the distinction has been preserved up to the present day, thus, Dan. brødre, Swed. brøder, døttre, fødder, høne, but sæde, nætter, læge, etc., Icel. bræðr, dætr, fætr, hæna, sæti, nætr, læknir, etc. The æ was sounded ē (as Germ. ä, in nähe, or a lengthened Engl. a, perhaps more protracted), this sound was still heard down to the time of Arni Magnusson (end of the 17th century), see Gramm. p. xxxv, col. 2; it is now sounded like Engl. long i (time), but this sound was unknown to the ancients, except in the interjection (= æ see below):—we can only guess at the sound of œ; judging from the analogy of æ, it may have been like Germ. höhe, Fr. feu, only more protracted; in a diphthongal form (like æ into  ) it would be oï, and indeed the word œli in its modern form auli (i.e. oïli) may be a relic of this.

2. v cannot be sounded before œ, being a u-sound; thus we have œðr, œða, from vaða, óðr; but before æ, being an a-sound, it is both sounded and written. For the umlaut see Gramm. p. xxix.

Ö (Ø)

This letter properly consists of two vowels, different in sound and in origin; an a-vowel, an ‘umlaut’ of a, and nearly related to it; and a u-vowel, nearly related to the letters o, u, and y: in modern Danish these two ö-sounds are still distinguished in pronunciation, the one being open almost like Engl. i before r, as in fir, the other closed like eu in French feu: Rask and Petersen, the founders of the philology of the Danish tongue, were the first to give separate symbols for these two sounds; the first they marked ö, the second ø (börn, høre). The modern Icel. knows only one sound, answering to the Danish ö; but that it was not so in old days may be proved from the vellums and from the grammarians. Thorodd marks the two sounds respectively by and ø. Most of the vellums are very loose in their spelling, marking at random o, , au, ꜹ, ø (oll, ll, aull, ꜹll): phonetically ø stands exactly in the same relation to œ, the umlaut of ó, as ö to ́, the umlaut of á, so that ø and ö are the short, œ and ́ respectively the corresponding long vowels; ø and œ, ö and ́ being two pairs of sounds, just as are o ó, u ú; cp. ‘Goðrøði’ and ‘góð rœði,’ Skálda, Thorodd: in very old vellums, e.g. the Rb. Cod. 1812, the ø is often marked eo, thus keomr = kømr, eoxn = øxn = yxn, geora = gjöra or gøra: in Norse vellums ø is often written œ, e.g. smœr = smjör, confounding the two sounds, ø and œ. A few good vellums keep the distinction in the main, not as Thorodd’s alphabet does, but generally by writing ey for ø (this must not be confounded with the diphthong ey); among those vellums are the Cod. Reg. of Sæm. Edda, the Cod. Acad. of the Hkr. (now lost), the Cod. Fris., the (lost) vellum of Rafns S. (see Bs. i. pref. lxix), although none of them strictly follows the rule; only a few Editions (e.g. Prof. Unger’s Edit, of the Hkr.) have tried to observe the distinction; most Editions print ö throughout. We shall now try to give a list of the chief words and forms which have the ø. The chief guide in doing this is twofold, the ey of the vellums and the change of ø into e or é, by which a triple form arises, ø, ey, and e, of which ø and ey, no doubt, are mere variations:

I. the ø is either,

1. the umlaut of o; in the plurals, sønir seynir senir, hnøtr hnetr, støðr steðr (sonr, hnot, stoð): in the compar. and superl., nørðri neyrðri nerðri, nørztr neyrztr nerztr, øfri efri, øfstr efstr (from norðr, of): in the subj., þørði þerði, þølði þeylði (Fms. viii. 380), møndi (from þora, þola, monu or munu), bjøggi beyggi, hjøggi heyggi (from búa, bjoggu, höggva, hjoggu): the presents, kømr, trøðr treyðr treðr, søfr sefr (from koma, troða, sofa): the prets., frøri freyri freri, gnøri gneyri gneri, søri seri, sløri sleri, róri reyri reri, kjøri keyri keri, snøri sneyri sneri, grøri greyri greri (see Gramm. p. xxiii): the words øðli eyðli eðli, øðla (a lizard) eyðla eðla, høllzti heylzti hellzti: in -røðr (Goðrøðr, see Thorodd), -frøðr -freyðr -freðr (Hallfrøðr Hallfreyðr Hallfreðr), hnøri hneyri hneri, øxn eyxn exn, køri keri (a probe), kjør (a choice) keyr ker, kjøptr keyptr keptr kjaptr: ørendi eyrendi erendi: the prefix particle, ør- eyr- er-: the words kjøt ket, smjør smér, mjøl mél (prop. køt, smør, møl), gørsemar gersemar, ørr and eyrr, a scar, Fms. viii. 275, v.l.; hrør and hreyr, heyrum and hørum (p. 261, col. 2).

2. in the case of roots in -vi or -vj, where both v and j struggle for the umlaut, the result is an ø; in this case even a radical a changes into ø (this was for the first time observed by the late Danish scholar Lyngbye), thus, gørva geyrva gera (from garvian), gørr geyrr gerr (= ready), gørvi gervi, gørsemi gersemi, øx eyx ex (Goth. aqwisi), sørvi seyrvi. This is esp. freq. in those roots which have g or k for the middle consonant, in which cases the root vowel, either a or i, changes into ø; as in the verbs sløkva, søkkva, støkkva, hrøkkva, kløkkva, sløngva, høggva, hnøggva, þrøngva; in the adjectives, døkkr, nøkviðr, gløggr, hnøggr, snøggr; similarly with the orthography ey for ø,—heygg (caedo), Am. 39; deyqva hramns, Skv. 2. 20; at kleycqvi Guðrun, Am. 58; klecqua, Akv. 24; hví er þér steyct ór landi, Hkv. Hjörv. 31; nú mun hón seyqvaz, Vsp. 62; seycstu nú gýgr (sink thou now!), Helr. 14; sleyngdi svá silfri, Am. 46; steyccr lúðr fyrir, Hkv. 2. 2; sýtir æ glæyggr við gjöfum, Hm. 48; gleyggr, Skv. 1. 7; gleggr, 291; neykðan (nudum), Am. 49; neycqviðr, Hm. 49; Beyggvir = böggvir, Ls. 45; røkvið and rekvið, Hkv. Hjörv. 35, Bugge (pref. ix); reykr = røkr, Fms. iv. 70: the word rekkja (a bed) is also spelt reykkja, and even rjukja, Art. (Ed. Kölbing) 64; vekka and vökvi. Phonetically connected with this change, but in a reverse order, is the change in the words nekkverr nökkurr nokkurr and eingi öngr öngvan, etc. In all the above instances the ey means ø, and is merely substituted for that sound, and is accordingly altogether different from the diphthong ey, see p. 114, col. 2, l. 15 sqq.

3. one may also assume an ø in the few instances where and jo, and jo and y interchange; in mjölk and mjolk (milk), mjok or mjök and mykill, þjökkr and þykkr, mjörkvi and myrkvi. This ø of the ancient tongue is the parent of the e in several modern words and forms, e.g. in the presents, sefr, kemr, treðr, heggr, sekkr, stekkr, hrekkr; in the preterites, greri, snéri, réri; the compar., efri, efstr, helztr; in gera, erindi, frerar, and freðinn: so also in the words két, mél, smér; and in inverse order, in nokkurr, in öngvir, öngvan, öngum, from einginn; cp. Dan. sen for sön.

The close phonetic relation between ø and y is shewn by the fact that g or k before an ø was sounded as an aspirate, thus, gøra, køt, køri, sounded gjöra, kjöt, kjöri, and more lately spelt with a j, analogous to gjæta, kjær, = gæta, kær, for the j in these words is not radical.

II. for the ö see Gramm. p. xxix, col. 2, and the introduction to letter A, p. 1. Runolf, in his Gramm. Island, of a.d. 1650, distinguishes between o longum (ó), o breve (o), and o brevissimum (ö).