OUTLINES OF GRAMMAB.

xxxv

(A. S. Beadu-); Dai (occurs in an Icel. colonist family from the Bri-
tish Isles in the loth century) probably from Dav (Davy]; Sebbi
and Ubbi occur on Swedish Runic stones; Helgi (old form Hlgi) from
Ha-leygr, Nj. ch. 94. Only a few instances in the Sagas bear directly
on this subject; one is the dream of earl Hakon (year 994) of his
son Erling's death ; ' n er Ulli daur,' qs. Erli or Erlingr; cp. also
the name of Snorri Goi from Snerrir, Eb. ch. 12. $0* Of a similar
kind are At-Ii, Goth, att-ila, Lat. patercuhts; Gam-li.

Compound Words.

Of these the Dictionary gives the best account; when the former
part is an uninflected root word a hyphen is usually printed between
the component parts, with a few exceptions, such as words com-
pounded with particles like afar-, all-, fjl-, full-, gagn-, etc.; and
some other words, as f-, go-, gull-, etc. Again, the Icel. has an
almost unlimited stock of compound words formed by means of the
genitive. Many of these are used both as compounds and as two |
separate words, and are therefore given under the head of the prin-
cipal word, e. g. barn with barns- and barna-: in these cases it depends
upon the genitive whether the alphabetical order is preserved or not;
this is mostly the case in words like btr, bts-bor, but not so in
ber, gen. bejar-; or in beini, beina-; baula, baulu-. As compounds
are made from both gen. plur. and sing, they are sometimes double,
e. g. under the head barn, both barns- and barna-. But chiefly are to
be noticed words with the u- umlaut, because a is the first and the
last letter in the alphabet; thus e. g. fur- is the compound form of
fair (father), and would if simple stand at the end of the letter,
whereas now it stands near the beginning, s. v. fair; as also bjarnar- \

under bjrn; bjarkar-under bjrk; still greater is the leap in com-
pounds from words such as alda, a wave, gen. oidu- ^p. 11) ; so also the
compounds from old (age), ond (soul), ork (arch), rn (eagle), xl
(shoulder), which are alitar-, andar-, arkar-, arnar-, axlar-; but these
words are few. Icel. printing, in editions of Sagas as well as in modern
books, has no fixed rule as to the spelling of such compound words,
and often connects them in hundreds of cases where they are evidently
separate; in old writers, e.g. in Mar. S., musterisfer,journey to the
temple,
14; frcistnistormr, storm of temptation, 433; uppstigningar-
star, place of ascension, 588 ; snubbanarorum, snubbing language,
567; uppsprettubrunnr, 27; stjrnubkarmerin, astronomers, 30;
spektargn, silence of wisdom, id.; umskr.r;irskrn, baptism of cir-
cumcision,
35 ; Austrvegskommgar, the kings of the East, id.; vistar-
veizluna, giving shelter, Mork. 67, etc.; and in mod. writers, e. g. in
the 4th hymn of the Passu-Slmar, trarsjnin, the eye of faith;
dreyralkir, brooks of blood; lausnargjald, 'lease-gild,' ransom; Kfs-
arnar, life veins; Arkargluggi, window of the Ark; hrygarskuggi,
the shadow of sorrow; slarbjarmi, the brightness of the sun ; hryg-
armyrkr, the darkness of grief; svalavain, the refreshing. water;
reiisproti, wrath's rod; svalalind, a refreshing well; hjartabl,
heart's blood, all spelt as one word, even without a hyphen between
them. Again, the old MSS. separate too much, or rather keep no
rule whatever. We have not thought of giving a full list of these and
similar words( for this would be impossible. From such words as
mar, barn, ftr, hond, etc. hundreds of similar compounds may easily
be formed, most of which are in a grammatical sense rather sentences
than single words ; but many are given, especially from old writers.
For a native these things are of little moment; but for the sake of
lexicography a more distinct and regular spelling is much needed.

REMARKS ON THE SPELLING IN VELLUM MSS.

A regular spelling has been adopted in most editions during the
last hundred years—before that time few editions had been issued :
this spelling was fixed by Icel. scholars of that time, and was chiefly
founded upon the average spelling in the vellums, partly upon a few
noted MSS. (e.g. the Arna-Magn. 132 folio, and 66 folio), and with
reference to the living Icel. language. But of late many of the
oldest MSS. and fragments have been carefully and exactly printed.
A few hints are therefore needed to guide the reader how in these
cases to use the Dictionary, which in the main holds to the normal
spelling. The spelling varies much, not only in MSS. of different
times, but in the same MS.; very few of them follow any fixed plan,
and the same word is differently spelt even in the same line; yet in
many particular instances the spelling is instructive, and even more
correct than the accepted orthography, and must not be left out of
sight by those who study the growth and history of the language.

A. In inflexions : I. vowels :—the MSS. use o and n as
well as e and i indiscriminately in declensions of nouns and verbs, the
oldest almost always o and e, as tungor, tongues ; oldor, waves; timom,
times; booot, kolloom, gorosk, etc.: e, i, as time, a time; elle, age;
fa.bei, father;
timenn, the time; boaer, fylger, etc.: most MSS. (the
later) prefer u, and so it has come into the normal spelling ; for the use
of e, see introduction to that letter (signif. B), p. 114 : in inflexions, -oil,

-orr, -or, -osta, -on, instead of-nil, -urr, -ur, -usta, -im (see pp. xxxii,
xxxiii); as also in dat.pl. with the article, timonom, hondonom; the
pret.tolo.fcj; kollo,i;oca.'a; kol\obom,vocavimus: a\so-endi,-enn,

-ell, instead of-indi, -inn, -ill. II. consonants:—the reflex, is in
very old MSS. spelt -sc (-zc or -sp~), but in the usual way -z, -zt, -szt.

B. In root syllables: I. vowels: 1. long and short
vowels are usually not distinguished, except in very few MSS., e. g.
Ann. Reg., which MS. is of a like interest for Icel. in this respect, as
the Ormulum for Early English. Later MSS. began to distinguish by
doubling the long vowels, aa — a, ij = i, oo = , w = , but mostly with-
out a fixed rule ; this way of spelling has remained in English, e. g.
Engl./oo/ = Icel. ft, blood = bl. At last the marking the long vowels
with an accent was resumed, as taught by Thorodd. 2. of
special letters, a. the spelling of varies very much ; the ancients
had a double sound (0 and co), but both were soon confounded, and
was spelt indiscriminately in a sixfold or eightfold fashion, o, w, an, av,
cO, 0 (born, barn, baurn, bavrn, boom, born), and was thus confounded
with several vowels, e. g. with the diphthong av, the o and 6, the ce and
œ, e. g. rar may be = rau red or r a row, log may be log a lowe or
log laws, lavg may be laug a bath or log laivs, huoll may be hsell a heel
or hll a hall, etc.; in print was used for about two hundred years,
till at the beginning of this century it was replaced by the present o,
which was probably borrowed from the German. p. the e and œ

were confounded, and in some few MSS. it is almost a rule, as the
Mork., the Njla (Arna-Magn. 468), the Kb. of Sm., and the frag-
ment Arna-Magn. 748, cp. e. g. the print of Baldrs Draumar in Sm.
Edda by Mbius, pp. 255, 256; thus tcki = tceki, seti = sti, reur —
rœur, beta —bœta, be = bee (a house), sekia — scekja, fela = fcela, mela
= maela, and vice versa; g, , instead of e, sgtti = setti, lli = elli, see
introduction to letter E, p. 113 ; œi = ei freq. In the east of Icel. the
ce and œ were, up to the beginning of the 18th century, sounded not =
Engl. long i as they are at present, but as Germ, or , Engl. a, with
a protracted sound : many pun.s referring to this provincialism art-
recorded by Jn lafsson, e. g. the ditty, mr s mcrin ( = mrin) Ijsa
i minni er,—the pun is in merr = n mare and mr = i maid being
sounded alike; Hann Bersi ininn i Be ! Hun er gengin 4 reur me
honum, see Jn lafsson, Essay on Icel. Orthography of the year 1756
(in MS.) The poet Stefn Olafsson, a native of the east of Icel. (died
1688), still rhymes brekr (i. e. braekr) and \ekr ( — stillai). It is likely
that the MSS. above named were written, if not composed, in the
east of Icel. In still earlier times this pronunciation was no doubt
universal, but not so six or seven hundred years ago. y- tne
Icel. (see p. xxix) confounded the two sounds (g) and œ (ad) ; yet for
a long time afterwards both characters g and ao were still used, but.
upside down, without any regard to etymology, till at last the Roman
took the place in writing of both g and ao. 8. the v and v were used
indiscriminately, e. g. tvngv = tungu, bvndv = bundu ; and, on the other
hand, ualld = valid, uera — vera, uit = vit, etc. t. the i served for i
andj (ior — jr) : ja is especially in very old MSS. often spelt ea, earn

— jam (cp. Thorodd in Sklda) : in old poems the j always serves as a -,
vowel in alliteration, which in mod. usage sounds harsh, though it may
be used ; but ia, io, etc. were, on the other hand, one syllable, and old
grammarians speak of /as a ' changeling,' being sometimes a vowel and
sometimes a consonant: it is likely that tiie pronunciation was similar
to ea in Engl. tears, fear, whereas in mod. Icel. usage j before a vowel
is sounded as Engl.^ before a vowel. f- 'll Norse MSS. ey is usually
spelt _y, hyra, yra, -— heyra, eyra, and is sounded thus in mod. Norse
dialects. ' TJ. many old Icel. MSS. confound^ and i in a few words and
forms, especially in the prepositions firir.ifir, = fyrir.yrir; the verbs skildi,
mindi (subj.), ikkir, --- skyldi, myndi, ykkir ; minni = mynni (ostium)
and minnask =- mynnask, ' to mouth,' to kiss; kirkja — kyrkja, cp. Scot.
kirk; before ngv, as singva = syngja to sing, Ingvi = Yngvi, lingva =
lyngva, etc.: mikill and mykill, mickle, much: the inflex. -indi and

-yndi. 9. the ey is used in some few MSS. instead of in such words
as seynir, seyni, = synir, syni; geyrva = grva. i. the o instead of
the later u in a few words, but only in very old MSS., as go = gu, goll
= gull, fogl = fugl, oxi = uxi, mon (the verb) = mun, cp. Engl. God, gold,
fowl, ox. K.
the p and œ are in very old MSS. spelt eo, e. g. keomr