Skip to main content

Full text of "Etymology of the principal Gaelic national names, personal names, surnames : to which is added a disquisition on Ptolemy's Geography of Scotland"

See other formats


v^ A 


1 ,>' 


,4 .^ J'?l 

QJocttcU Hniueraitg Siibrarg 

3tt)ara, 'Ntm fork 






Cornell University Library 
PB 1583.M1109 

3 1924 026 841 001 

Cornell University 

The original of this book is in 
the Cornell University Library. 

There are no known copyright restrictions in 
the United States on the use of the text. 










The following Outlines of Gaelic Etymology originally formed 
part of, and was bound up with, the first edition of the Gaelic 
Etymological Dictionary by the late Dr MacBain. The publisher, 
now thinking that there are students of the Language who might 
wish to have the "Outlines" in a separate and handy form, is 
here publishing them. • 

The Supplement, the words and letters in square brackets, 
and a few slight changes from the original are the work of the 
Rev. Dr George Henderson, Lecturer in Celtic Languages and 
Literature in the University of Glasgow, who found it necessary 
to abandon his intention of seeing the Gaelic Etymological 
Dictionary through the press, after reaching the sixteenth page 
of these "Outlines." 


•Gaelic belongs to the Celtic group of languages, and the Celtic is 
itself a branch of the Indo-European^OT Aryan family of speech ; 
for it has been found that the languages of Europe (with the 
■exception of Turkish, Hungarian, Basque, and Ugro-Finnish), and 
those of Asia from the Caucasus to Ceylon,^ resemble each other 
in grammar and vocabulary to such an extent that they must all 
be considered as descended from one parent or original tongue. 
This parent tongue is variously called the Aryan, Indo-European, 
Indo-Germanic, and even the Indo-Celtic language. It was 
«poken, it is believed, some three thousand years B.C. in ancient 
Sarmatia or South Eussia ; and from this as centre ^ the speakers 
of the Aryan tongue, which even then showed dialectal differ- 
ences, radiated east, west, north and soiith to the various countries 
now occupied by the descendant languages. The civilization of 
the primitive Aryans appears to have been an earlier and more 
nomadic form of that presented to us by the Celtic tribe of the 
Helvetii in Caesar's time. Here a- number of village communities, 
weary of the work of agriculture, or led by the desire of better 
^soil, cut their crops, pulled down their lightly built houses and 
huts, packed child and chattel on the waggons with their teams of 
•oxen, and sought their fortune in a distant land. In this way 
the Celts and the Italians parted from the old Aryan home to 
move up the Danube, the former settling on the Ehine and the 
latter on the Gulf of Venice. The other races went their several 
ways — the Indians and Iranians eastward across the steppes, the 
Teutons went to the north-west, and the Hellenes to the south. 

The Aryan or Indo-European languages fall into six leading 
groups (leaving Albanian and Armenian oat of account), thus : — 

I. Indo-Iranian or Arian, divisible into two branches : 

(a J Indian branch, including Sanskrit, now dead, but dating 
in its literature to at least 1000 B.C., and the descendant 
, modern (dialects or) languages, such as Hindustani, 
Bengali, and Mahratti. 

^ ^ See Supplement to Outlines of Gaelic Etyntology. 



(b) Iranian branch, which comprises Zend or Old Bactrian 
(circ. 1000 B.C.), Old Persian and Modern Persian. 

II. Greek or Hellenic, inclusive of ancient and modern Greek 

(from Homer in 800 B.C. onwards). Ancient Greek was 
divided traditionally into three dialects — Ionic (with Attic or 
literary Greek), Doric, and ^olic. , 

III. Italic, divided in early times into two main gronps — th& 
Latin and the Umbro-Oscan. From Latin are descended 
Italian, French, Spanish, Portuguese, Ehoeto-romanic and 
Eoumanian, called generally the Eomance languages. 

IV. Celtic, of which anon, 

Y. Teutonic, which includes three groups — (a) East Teutonic or 
Gothic (fourth cent, a.d.) ; (bj North Teutonic or Scandi- 
navian, inclusive of Old Norse and the modern languages, 
called Icelandic, Norwegian, Swedish, and Danish ; and (c)- 
West Teutonic, which divides again into High German 
(whence modern German), the Old High German being a 
language contemporary with Old Irish, and Low German, 
which includes Old Saxon, Anglo-Saxon, English, Dutch, and 

VI. Balto-Slavonic or Letto-Slavonic, which includes Lithu- 
anian, dating from the seventeenth century, yet showing 
remarkable traces of antiquity, Lettic, Old Prussian of the 
fifteenth and sixteenth centuries, now extinct, Old Bulgarian 
or Church Slavonic, into which the Bible was translated in 
the ninth century, and the Slavonic modem languages of 
Kussia, etc. , 

These six groups cannot, save probabiy in the case of Latin* 
and Celtic, be drawn closer together in a genealogical way. 
Eadiating as they did from a common centre, the adjacent groups 
are more like one another than those further off. The European 
languages, inclusive of Armenian, present the three primitive 
vowels a, e, o intact, while the Indo-Iranian group coalesces them 
all into the sound a. Again the Asiatic languages join with the 
Balto-Slavonic in changing Aryan palatal k into a sibilant sound. 
Similarly two or three other groups may be found with common 
peculiarities (e.g., Greek, Latin, and Celtic with oi or i in the nom. 
pi. maso. of the o- declension). Latin and Celtic, further, show 
intimate relations in having in common an i in the gen. sing, of 
the 0- declension (originally a locative), -Hon- verbal nouns, a 
future in b, and the passive in -?•. 

' See Sutpplcmcnt to OvMines of Oaelic Etymology. 


The Celtic group now comprises five living languages ; in the 
18th century there were six, when Cornish still lived. These six 
Celtic languages are grouped again into two branches, which may 
be named the Beittonic and the Gadelic. The former includes 
the Welsh, Cornish, and Breton ; the Gadelic comprises Irish, 
Manx, and (Scottish) Gaelic. The main difiference between these 
two branches of the Celtic group consists in this : the velar 
guttural of the Aryan parent tongue, which we represent here by 
the symbol q, when labialised, that is when the sound w m u 
attaches itself to it, becomes in Brittonic a simple p and in 
Gadelic a c {k. Ogam qvi,). Thus the Welsh for "iive" is jtump, 
Cornish pymp, and Breton pemp, Gaulish pempe. whereas the 
Gaelic is c&ig, Manx queig, and Irish cMg : the corresponding 
Latin form is quinqiue. Professor Rhys has hence called the two 
branches of the Celtic the P group and the Q group (from Ogmic 
Q'M=' Gaelic c). The distinction into P and Q groups existed 
before the Christian era, for the Gauls of Caesar's time belonged 
mainly, if not altogether, to the P group : such distinctive forms 
as Gaulish petor, four (Welsh pedwar, Gaelic ceithir), epos, horse 
(Welsh ehol, Gaelic each), and pempe, five, already noted, with 
some others, prove this amply. At the beginning of the 
Christian era the Celtic languages were distributed much as 
follows : Gaulish, spoken in France and Spain, but fast dying 
before the provincial Latin (and disappearing finally in the fifth 
century of our era) ; Gallo-Bkitish or Brittonic, spoken in 
Britain by the conquering Gaulish tribes ; Pictish, belonging to 
the Gallo-Brittonic or P group, and spoken in Scotland and, 
possibly, in northern England ; and Gadelic, spoken in Ireland 
and perhaps on the West Coast of Scotland and in the Isles. The 
etymology of the national names will be seen in Appendix A. 
Our results may be summed in a tabular form thus : — 

i Irish 
( Dialects in Spain and Gaul (1)* 


P Group 

Gallo-Brittonic [ Brittonic . . { Cornish 

I (Welsh 

Gaulish — various 

There are no literary remains of the Gaulish language existent ; 
but a vast mass of personal and place names have been handed 

^ ^ See Supplement to Outlines of Gaelic Etymology. 


down, and also a few words of the ordinary speech have been 
recorded by the Classical writers.^ The language of Brittany came 
from Britain in the fifth and sixth centuries, and it may have 
found remains in Brittany of the kindred Gaulish tongue. The 
Brittonic languages — Welsh, Cornish, and Breton — appear first in 
glosses as early as the eighth century. These glosses are 
marginal or super-linear translations into Celtic of words or 
phrases in the Latin texts contained in the MSS. so "glossed." 
The period of the glosses is known as the "Old" stage of the 
languages — Old Breton, Old Cornish, Old Welsh. Real literary 
works do not occur till the "Middle" period of these tongues, 
■commencing with the twelfth century and ending with the six- 
teenth. Thereafter we have Modern or New Breton'^ and Welsh 
as the case may be. In this work, JSTew Breton and New Welsh 
are denoted simply by Breton and Welsh without any qualifying 

The Gaelic languages — Irish, Manx, and Scottish Gaelic — 
have a much closer connection with one another than the 
Brittonic languages. Till the Reformation and, indeed, for a 
century or more thereafter, the Irish and Scottish Gaelic had a 
common literary language, though the spoken tongues had 
diverged considerably, a divergence which can be traced even in 
the oldest of our Gaelic documents — the Book of Deer. In the 
eighteenth century Scottish Gaelic broke completely with the 
Irish and began a literary career of its own with a literary dialect 
that could be understood easily all over the Highlands and Isles. 
Manx is closely allied to Scottish Gaelic as it is to the Irish ; it 
is, so far, a remnant of the Gaelic of the Kingdom of the Isles. 

The oldest monuments of Gaxielic literature are the Ogam 
inscriptions, which were cut on the stones marking the graves of 
men of the Gaelic race. They are found in South Ireland, Wales 
and Eastern Pictland as far as the Shetland Isles, and belong mostly 
to the fifth, sixth, seventh, and eighth centuries. The alphabet, 
which is formed on a proto-telegraphic system by so many strokes 
for each letter above, through, or below a stem line, is as 
follows^ :— 

I m i l IN I mil ' " '" "" '"" 

b, 1, f, s, n; h, d, t, c, q ; 

/ // /// //// ///// l ^-W-Wt-HW • 

">< g. °g. z. r ; a, o, u, e, i. 

* ' ^ See Su^lemcnt to Outlines of Gaelic Etymology, 


Examples of Ogam inscriptions are : — 
Sagramni maqi Cunotami 

" (The stone) of Sagramnos son of Cunotamus." 
Maqi Deceddas avi Toranias 
" Of the son of Deces 0' Toranis." 
Cunanettas m[aqi] mucoi Nettasegamonas 
" Of Cunanes son of the son of Kettasegamon." 
Tria maqa Mailagni 
"Of the three sons of Maolan." 

These examples show th^t the state of declensional inflection was 
as high as that of contemporary Latin. The genitives in * belong 
to the o declension ; the i, as in Old Irish, is not taken yet into 
the preceding syllable (maqi has not become maic). The genitives 
OS and as belong to the consonantal declension, and the hesitation 
between a and o is interesting, for the later language presents 
the same phenomenon — the o in unaccented syllables being 
dulled to a. The Ogam language seems to have been a preserved 
literary language ; its inflections were antique compared to the 
spoken language, and Old Irish, so near it in time as almost to be 
contemporary, is vastly changed and decayed compared to it. 
Irish is divided into the following four leading periods : — 

I. Old Irish : from' about 800 to 1000 a.d. This is the period 

of the glosses and marginal comments on MSS. Besides 
some scraps of poetry and prose entered on MS. margins, 
there is the Book of Armagh (tenth century), which contains 
continuovis Old Irish narrative. ^ 

II. Eakly Irish, or Early Middle Irish : from 1000 to 1200 a.d. 

— practically the period of Irish independence after the 
supersession of the Danes at Clontarf and before the English 
conquest. The two great MSS. of Lehor na h-uidre, the 
Book of the Dun Cow, and the Book of Leinster mark this 
period. Many documents, such as Cormac's Glossary, claimed 
for the earlier period, are, on account of their appearance in 
later MSS., considered in this work to belong to this period. 

III. Middle Irish : from 1200 to 1550 (and in the case of the 
Four Masters and O'Clery even to the seventeenth century in 
many instances). The chief MSS. here are the Yellow Book of 
Lecan, the Book of Ballimote, the Leahar Breac or Speckled 
Book, and the Book of Lismore. 

IV' Modern, or New Irish, here called Irish : from 1550 to the 
present time. 

' See Supplement to Outlines of Gaelic Etymology, 


As already said, the literary language of Ireland and Scotland 
remained the same till about 1700, with, however, here and there 
an outburst of independence. The oldest document of Scottish 
Gaelic is the Book of Deer, a MS. which contains half a dozen 
entries in Gaelic of grants of land made to the monastery of Deer. 
The entries belong to the eleventh and twelfth centuries, the most 
important being the first — the Legend of Deer, extending to 19 
lines of continuous prose. These entries form what we call Old 
Gaelic, but the language is Early Irish of an advanced or phoneti- 
cally decayed kind. The next document is the Book of the Dean 
of Lismore, written about 1512 in phonetic Gaelic, so that we may 
take it as representing the Scottish vernacular of the time in 
inflexion and pronunciation. It differs considerably from the 
contemporary late Middle Irish ; it is more phonetically decayed. 
We call it here Middle Gaelic, a term which also includes the 
MSS. of the M'Vurich seanchaidhean. The Femaig MSS.,!" written 
about 1688, is also phonetic in its spelling, and forms a valuable 
link in the chain of Scottish Gaelic phonetics from the Book of 
Deer till now. The term Gaelic means Modem Gaelic. 

Scottish Gaelic is written on the orthographic lines of Modem 
Irish, which in its turn represents the orthography of Old Irish. 
The greatest departure from ancient methods consists in the 
insistence now upon the rule of " Broad to broad and small to 
small." That is to say, a consonant must be flanked by vowels of 
the same quality, the "broad" being a, o, u, and the "small" e 
and i. Gaelic itself has fallen much away from the inflexional 
fulness of Old Irish. Practically there are only two cases — nom. 
and gen. : the dative is confined to the singular of feminine nouns 
(a-declension) and to the plural of a few words as laid down in the 
grammars but not practised in speech. The rich verbal inflexion 
of the old language is extremely poorly represented by the 
impersonal and unchanging forms of the two tenses — only two — 
that remain in the indicative mood. Aspiration, which affects all 
consonants now, (though unmarked for I, n, r), has come to play 
the part of inflection largely ; this is especially the case with the 
article, noun, and adjective. Eclipsis by n is practically un- 
known ; but phonetic decay is evidenced everywhere in the loss of 
inflection and the uniformising of declension and conjugation. 

There are two main Dialects of Gaelic, and these again have 
many sub-dialects. The two leading Dialects are known as the 
Northern and Southern Dialects. The boundary between them 
is described as passing up the Firth of Lorn to Loch Leven, and 
then across from Ballachulish to the Grampians, and thence along 

'" See Swpplemcnt to Outlines of Gaelic Etymology. 


that range. The Sovithern Dialect is more Irish than the 
Nbrthem, and it has also adhered to the inflections better {e.g., 
the dual case still exists in feminine a nouns).!'- The crucial dis- 
tinction consists in the different way in which the Dialects deal 
with e derived from compensatory lengthening j'^ in the South it is 
€u, in the North ia {e.g., few against ^a/*, breug against hriag, &c.) 
The sound of ao differs materially in the two Dialects, the 
Southern having the sound opener than the Northern Dialect.'^ 
The Southern Dialect is practically the literary language. 

Modern Gaelic has far more borrowed words than Irish at any 
stage of its existence. The languages borrowed from have been 
mainly English (Scottish) and Norse. Nearly all the loan-words 
taken directly from Latin belong to the Middle or Old period of 
(jlaelic and Irish ; and they belong to the domain of the Church 
and the learned and other secular work in which the monks and 
the rest of the clergy engaged. Many Latin words, too, have been 
borrowed from the English, which, in its turn, borrowed them 
■often from French, (such as prls, cunntas, cidrt, spbrs, ifec). Latin 
words borrowed directly into English and passed into Gaelic are 
few, such as post, plasd, peur, &c. From native English and from 
Lowland Scots a great vocabulary has been borrowed. In regard to 
Scots, many words of French origin have come into Gaelic through 
it. At times it is difficult to decide whether the Teutonic word 
_,was borrowed from Scottish (English) or from Norse. The con- 
tributions from the Norse mostly belong to the sea ; in fact, most 
■of the Gaelic shipping terms are Norse. 

Under the heading of Phonetics we deal with the sounds of 
the language — the vowels, semi-vowels, and consonants, separately 
and in their inter-action upon one another. 

§ 1. Alphabet. 

The Gaelic alphabet consists of eighteen letters, viz., a, b, c, d, 
€, f, g, Ji, 1, I, m, n, o, p, r, s, t, and u. Irish, Old and New, have 
the same letters as the Gaelic. As this number of letters in no 
way adequately represents the sounds, signs and combinations are 

Firstly, the long vowels are denoted by a grave accent : ct, i, 
ic, e, d, the latter two having also the forms e, 6, to denote sounds 
analogous to those in English vein, boa?: Whereas &, I, u, which 
have only one sound, represent corresponding Indo-European 
sounds (a, t, u), none of the long . sounds of e or o represent - 
simple corresponding I.E. sound. 

11 12 J3 ggg Supplement to Outlines of Gaelic Etymology. 


The Gaelic vowels are divided into two classes — broad and 
small. The broad vowels are a, o, u ; the small, e, i. The Gaeli& 
diphthongs^* represent (1) simple sounda, (2) real diphthong sounds, 
or (3) modification of the consonants and carrying out of the law 
of "broad to broad and small to small." They are as follows : — 


ao, [aw]* 



ei, eo, eii, eb 



to, iu, iu 






, ui 


Here ea, ei, eu represent 0. Ir. e, d, and are practically simple 
sounds, as certainly is ao. The forms ia, ua are genuine diph- 
thongs, as are usually the long vowel combinations. The rest 
may be diphthongs, or may be a trick of spelling, as in the 
word fios (0. Ir. fis), where the o shows that the s has its normal 
sound, and not that of E. sh, as fls would imply. 

Triphthongs occur in the course of inflection, and in the case 
of ao otherwise. These are — aoi, eoi, iai, iui, uai, ebi, int. 

The consonants are classified in accordance with the position 
of the organs of speech concerned in their utterance : — 

I. Liquids. — The liquids are I and r, with the nasals n and vi. 
In writing, m only is "aspirated," becoming to the eye mh, to the 
ear a v with nasal influence on the contiguous vowels. The other 
liquids, /, n, and r, are really aspirated in positions requiring 
aspiration, though no 7i is attached to show it.^^ There is, however, 
only a slight change of sound made in these letters by the aspira- 
tion — a more^'' voiced sound being given them in the aspirating 

II. Mutes and Explosives. — These all suffer aspiration when 
intervocalic. They are classified as follows : — 

Tenues. Medire. Aspirates. 

Labials /) b ph, Ih 

Dentals t d th, dh 

Gutturals c g ch, gh 

The dentals d and t become spirants i" when in contact with, or 
flanked by, the "small" vowels e and i. The other mutes are not 
affected by such contact.i^ The aspirate sounds are^^/j=/; 
bh = V, th = h, dh and gh before e, i = y, ch = German and Scotch cA. 

* Dialectal, before U, nn, rah, bh, though not in the script. 

14 15 16 17 18 gge Supplement to Outlhies of Gaelio Etynwloyy. 


III. The Spirants. — These, outside the above spirant-made 
mvites, are / and s. The sound [resembling E.] sh is represented 
by s flanked with "small" vowels. The aspirate forms of these 
are — jh ( = the Greek open breathing or nothing practically),{=h). 

Celtic Alj^habet. 

The Celtic alphabet, as deduced from the Neo-Ccltic dialects, 
checked by Gaulish, possessed the following sounds : — 

I. VowEM : — 

Short — i, V, e, o, a 

Long — I ( = T, e), u, e ( = ei), o { = mi), u ( = o, «) 

Diphthongs — ei, oi, cd, eu, ou, cm 

II. Liquids — r, I, m, n 

III. Spieakts — {h), s, j, V 

IV. Explosives : — Tenues. Mediic. 

Labials — b 

Dentals t d 

Gutturals 1;, lev, (p) y, gv {0) 

It has to be noted that Indo-European p initial and intervocalic 
is lost in Celtic.^8 Before another consonant, it manifests its former 
presence by certain results which still remain. Thus I. E. septn, 
is G. seachd, supno-s becomes suan. 

Indo-European Alphabet. 

By a comparison of the six Indo-European or Aryan language 
groups, the sounds possessed by the parent tongue may be 
inferred. The following is the form of the I. E. alphabet which is 
used in the present work : — 

I. Vowels : Short — i, u, e, o, a, 9 

Long — ^, u, e, 0, a 
Diphthongs — ei, oi, ai, eii, on, au 
ei, oi, di, eit, ou, du 

II. Semi-vowels : i, u, represented in this work always by 

j, V. See the spirants. 

III. Coxsonant-vowels : ?•, I, 111, n, f, I, 111, n 

IV. Liquids and Nasals : r, I, m, n 

V. Spirants -.j, v, s, z 

" See Sujyphment to Outlines of Oaelic Etymolo'jy. 


VI. Explosives^" : — Teuues. Media;. Aspirates. 

Labial „. p h ph, hli 

Dental t d th, dh 

Palatal h g kh, gh 

Velar q g 'jh, gh 

§ 2. Vowel Modification, 
In Gaelic the vowel or vowel combination of a syllable may 
nndcrgo "mutation" (German imilaut) in the course of inflection 
or word-building. This mutation is caused by the influence 
exerted backward by the vowel of the next syllable now or previ- 
ously existent. There are three classes of mutation in Gaelic 
caused either by a following (!) e ov i, (2) a or o, or (3) u. 
Jhitation hy " e" or "i." 
a becomes (1) ai : eat, gen. cait, damh, g. daimh. 

(2) oi (with double liquids usually) : dall, pi. doill, 

clann, g. cloinne. 
(•3) ui (with liquids) ; hall, pi. huill, allt, g. uillt. 
Also where Irish shows o : halg, 0. Jr. hole, 
pi. huilg ; so dag, fait, gal, fuil, car. 
(■i) i : mac, g. mic. Dialectally ai becomes ei, 
especially with liquids, and in ordinary G. 
eile represents 0. Ir. aile ; so seileach, too. 
o becomes (1) oi : sgoltadh, sgoilte. 

(2) ui : bonn, g. buinn, 2)ost, g. /)!(«'«<. 
u becomes ui : dubh, conip. didhhe. 
e becomes ei : beir for *here, catch thou. 
a, 6, u become ai, oi, ui : laimhe, oige, diiin. 
eo, iu, ua become triphthongs ; [the digraph ao + i form's a 

ea becomes (1) ei : each, g. eich. 

(2) i : ceann, g. cinn ; the usual mutation. 
en, with liqviids, becomes ebi : beul, g. hebil. It sometimes 

becomes ao : eudann, aodann. 
ia is restored to ei : fiadh, g. feidh : irregularly— j^ar, crooked, 

comp.j^aiVe, biadh, g. bidh, [Dial. b(iidk, beidh, bi-idh.] 
io becomes i :fionn, g.finn. 

Mutation hy " o" or " a." 
becomes a, a mutation of principal syllables rare in Irish : 

caf, Ir. cos, original *coxa ; cadcd for codal. 
u becomes o : sruth, g. srotha ; nuadh, nodhct. 
e becomes ea : cearc from *cerca. 

-" See Siqoplcment to Ouilincs of Gnelio Etymatoyy. 


i becomes ea : fear from *viro-s. 

di becomes ia : the stem feidh becomes fiadh in the nom. 

I becomes \o : flor from *vvro-s. 

Mutation hy "m." 
A succeeding u affects only i or e ; it is a mutation which does 
not now operate. Thus ftodh comes from *vidit- (0. Ir. fid) ; bior 
from *beru (0. Ir. bir) ; sliochd from slektu- ; cionn from the dat. 
*cennu, from *cennd. 

§ 3. Indo-European and Gaelic Vowels. 

The representation in Gaelic of the I. E. vowels is very com- 
plicated owing to the principles of mutation discvissed above. 

I. E. i. 
(!) Gaelic i, 0. Ir. i, W. y. 

bith, world, 0. Ir. bith, W. byd, Br. bed : *bitu-s, root gi. So 
it/i, fidir, nigh, fir (gen. and pi. of fear), as also nid from nead, 

(2) G. ea, 0. Ir. e. 

beatlia, life, 0. Ir. bethu : *bitus, stem *bibdt-, root gi. So 
eadh, it,.fear, geamhradh, meanhh, nead, seas, seasg, sleamhmwn, 

(3) G. io, 0. Ir. i. 

G. iiodh, wood, 0. Ir. fid, W. gwydd, Br. gwez : *vidu-. So 
fios, iodh-. The io of fionn, 0. Ir ^mc? is due to the liquid 
and medial mute, which together always preserve the i and 
even develop it from an original n or en (nb,. nd, ng). 

(4) G., 0. Ir. iu. 

This is a mutation by u : fliuali, wet, from *vliqvr ; tiugh, 

I. E. u. 

(1) G., 0. Ir. u, W. w (o). 

G., 0. Ir. sruth, stream, W. frwd : *srutu-s. So bun, dubh, 
guth, imM, WMsach, dug, smug, tulach. 
Here add G. ui : cluinn, hdbh, uisge. 

(2) G., 0. Ir. o. 

bonn, bottom, 0. Ir. bond, W. hon, *bundo-s. So bothan, con, 
dogs', do-, so-, domhan, dorus, torn, os, trod. 
I. E. e. 

(1) G., 0. Ir. e, W. e. 

Simple e is rare in G. : leth, side, 0. Ir. leth, W. lied, *letos. 
So teth, hot. 


(2) G. ea, 0. Ir. e. 

G. each, horse, 0. Ir. ech, W. ebol, Lat. equus. So numeroxxs 
words — eadh, space, hean, heart, cearc, ceart, dean, dearg, 
deas, fearg, geal, geas, meaclJion, meanmna, meas, neart, reachd, 
seacli, seaclid, sean, searg, teach, teas, treabh. 

(3) G. ei, 0. Ir. e. 

G. beir, take, 0. Ir. herim, W. adfer, Lat. fero. So beil 
{meil), ceil, ceirtle, ceithir, creid, deich, deis, [Dial.] ready, 
meirbh, seinn, teich, teine. 

(4) G., 0. Ir. i. 

G., 0. Ir. fine, tribe, root ven, 0. H. G. wini, kg. S. wine, 
friend. So cineal, gin, ite, mil, misg, sinnsear, tigh, tighearna. 

(5) G. io, 0. Ir. i. 

G. bior, spit, 0. Ir. hir, W. her, Lat. veru. So iol-, sliochd^ 
smior, biolaire, ciomach, tioram. 

(6) G. ui ill ruith, ruinn = rinn (bis), ruighinn and righinn : (Cf. 

roinn, [Dial.] did, for rinn ; ruigheachd). So trusdair, 

(7) Compensatory long vowels in G. and 0. Ir. These arige from 

loss of one consonant before another, one of frhich must be a 

a. ent becoifies G. eiid, 0. Ir. 'et. G. ceud, first, O. Ir. cit, W. 

cynt. So seud, journey. Similarly '''enk ; G. eug, death, 
0. Ir. ic ; *brenkd, G. breug, lie, 0. Ir. brec, ; *enH, 
G. eitchd, E. Ir. e'cht (Cf. creuchd, *crempt- ?) ; *cents6 ; G. 
ceuB, crucify. Parallel to these forms in ent, enk are 
those in nt, nk, such as ceud, one hundred, 0. Ir. cet, 
W. cavt, Lat. centum (so deud, eug, geug). 

b. ebl : in G. neul, cloud, 0. Ir. nel, W. niwl. 
egr : in G. feur, grass, 0. Ir. fdr, W. gwair. 
egn : in G. feun, 0. Ir. fen : *vegno-s. 

etl : in G. sgeul, 0. Ir. seel, W. chwedl. 
etn : in G. ev/n, 0. Ir. 6n, W. edn. 

c. G. eadar and thig show short vowels for original *enter and 

enk. This is due to sentence accent in the case of eadar 
and to the word accent in the case of thig or to both. 

For ceum, leiim, etc., see under it. 

I. E. o. 
(1) G., Ir. o. 

G. CO-, comh-, with, 0. Ir. co-, com-, W. cy-, cyf-, *kom- ; so ro- 
( = Lat. pro), fo (='Gr. iVd), nochd, naked, night, ochd, mol, 
iodhar, gon, gort, roth. 


(2) G., 0. Ir. u, ui. 

G., 0. Ir. muir, sea, W. mor, Br. mor, from *m,ori. So druim 
(*dros-m,en), guidlie, guil, guin, sguir, suidhe, uidlie, tdleann, 
uireemi, gu, to, cm-, fu-, fwr- {fur = *vor). 

(3> 3. a, 0. Ir. o. 

G. cas, foot, 0. Ir. cos, W. coes, *coxd. So am/j, fta^^r, call, fait, 
yart, gar, calltuinn. So, too, compounds. With con as in 
zagainn, cadal, cagar, caisg, as against coguis (0. Ir. concubus), 
with its M §oimd terminal. 

:^* compensatory long vowels. 

G. dual, look of hair, *doglo-. Got. <a^^, Eng. to«7. So bl 
(*potlo-), buain, {*bog-ni- or *bongni-), cluain, cuan, bruan, 
srdn, cdmh-. 

I. E. a. 
{!) G. a, ai, 0. Ir. a, W. a. 

G., 0. Ir. can, sing, W. ea»a, Lat. cano. So many words, 
such as abhainn, ad-, agh, air, altrimn, anail, anam, cac, 
damh, gad, mac, maide, marc, nathair, salann, &c. 

(2) G. ct before rd, m, in. 

See ard, bard, barr, cam, sghird, cam, am, mam, 

(3) G. i. 

In two cases only : mac, g. mic ; sile [Dial, for seile\, saliva, 
0. Ir. ■■ 

(4) G. u, ui. 

This happens in contact with liquids. The prep, air becomes 
ur-, uir-, urchar, uireasbhuidh. So muigh from *rrMgesi. 
Common in oblique cases : allt, g. uillt, ball, buill, &c. 

{5) G. ea, ei for e. 

G. seileach, willow, E. Ir. sail, W. lielyg, Lat. salix. So 
eaUuinn, eile, eir- for air-, eilean, [Dial.] training, deigh, ice. 

(6) G. oi. 

This change of I. E. a into Gaelic oi is due mostly to a liquid 
followed by a "small" vowel. 

G. oil, rear, E. Ir. aillm, Lat. alo. So oir for air-, coileach, 
goir, troigh, coire, loinn, &o., and goid, oide. 

(7) Compensatory lengthenings in G. 

a. As a, ai : 

G. dail, meeting, 0. Ir. ddl, W. dadl, where -ado- is the 
original combination, -agr- appears in naire, sar, d,r. 


b. As eii, ao, ia : 

It has been seen that ceud, hundred, corresponds to W. 
cant, Lat. centum. The Celtic, in these cases, is regarded 
as having been nt, nh, (*hnto-n). See under n. 
, An undoubted case of a landing by compensation into 
eu { = 'e) is rfew, tear, 0. Ir. ddr, 0. W. dacr, I. E. ddkru. 
Prof. Strachan has extended this analogy to words like 
meiir, breun, leine, sc/eun, meanan. The case of deur 
seems rather to be an anomaly.^'- 

IE. 5. 
This is the I. E. " indefinite'' vowel, appearing in Celtic as a, 
in the Asiatic groups as i, and generally as a in Europe (Greek 
showing also e). Henry denotes it by a, a more convenient form 
than Brugmann's 9. Some philologists refuse to recognise it. 

G. athair, father, 0. Ir. athir, I. E. pster-, Gr. Trar-qp, Skr. 
It is common in unaccented syllables, as G. anail, breath, 
W. anadl, *an^-tla, Gr. avefios. In the case of syllables with 
liquids it is difficult to decide whether we have to deal with a, 9, 
or a liquid vowel ; as in G. ball, member, *bhal-no-, root bhs>l, 
whence Gr. <^aAAds, Eng. hole. 

I. E. Long Vowels. 

I. E. I and u are so intimately bound with ei and en, (on) 
that it is difficult to say often whether we have to deal with the 
simple vowel or the diphthong as the original. For i see li, sin, 
sglth, hrigh ; for u, see cid, duil, element, dim, cliii, much, miiin, 
run, iir. The W. in both cases (i., u) show simple i. 

I. E. e appears in Celtic as I, G. % : as in G. flor {fir), true, 
0. Ir. fir, W. and Br. gwir, Lat. verus. So Iwn, mial (mlol), mlo& 
righ, stth, s\ol, sior, tlr, sniomh. 

I. E. and a appear both as a in the Celtic languages — 
Gadelic a, W. aw, Br. eii.. For o, see blatk, gnath, lar, dan, snath. 
For a, see ban, hrhthair, cnaimh, car, clar, daimh, faidh, gair. 
mathair, sath, tamh. But rbin, ron, nbs, main, all from a? b in' 
finals, etc., may equal m : *svesor = 0. Ir. siur, fiur, Med. Ir. siiir. 

I. E. Diphthongs. 
I. E. ei (ejf) appears in G. in two forms — as e'i and ia. Thus — ■ 
a. G. ei, 0. Ir. ei, W. wy, Br. oe, oa. See feith, gMl, meith, 
r6idh, sdid, smiid?''^ 

^' -- See Supplement to Outlines of Oaejic Etymology. 


h. G. ia, 0. Ir. ia. This is due to the influence of a succeed- 
ing broad vowel. See cia, ciall, cliathach, criathar, Hadh, 
fiamiis, giall, iarunn, liagh, riadh, riar, sgiath, sliabh. 
Consider these — feuch, lean, gle, and, possibly, geadh. 
I. E. oi {6} ?). This consistently appears in G. as ao long, 
0. Ir. di, 6i, later oe, ae, {6e, de), W., Br. u. See caonih, claon, 
fraoch, gaoth, gaol, laogh, maoin, maoth, taobh. 

I. E. ai can with difficulty be dift'erentiated from oi ; certainly 
not on Celtic ground, nor, indeed, outside Greek and Latin. The 
following are real cases : G. doii, caoch, saothair, taois. 

I. E. eu and ou are also confused together in the modern Celtic 
languages. They both appear as either G. ua or 6. 
a. G. ua, 0. Ir. 4a, W., Br. u. 

G. hvAxidh, victory, 0. Ir. huaid, W. hud, Gallo-British 
Boudicca, "Victoria." See also buachaill, cluai, luath 
ruadh, ruathar, truagJi, tunth, uasal. 
h. G. 6 ; as hbidheach from huaidli, trbcair from truagh, 
lochran, cos for cuas. ' 

I. E. au^^ appears in G. as o or ua, much as do eu, ou. Thus — ■ 
G. go, a lie, 0. Ir. gd, gdu, W. gau, Br. gaou. Also high, virgin, 
from augi-, fuachd, uaigtieach. 

§ 4. I. E. Semi- Vowels and Consonant Vowels. 
The semi-vowels are denoted by Brugmann as i and ^l, by 
Henry as y and w ; and these forms are used by them " not m'erely 
for intervocalic semi-vowels but also for the diphthongs which we 
have printed as ei, oi, ai, eu, ou, au, which Henry, for instance, 
prints as ey, ew, etc. In this work Fick is followed in the forms 
of the diphthongs, and also, where necessary, in his signs for the 
semi- vowels, viz., y and v, with j and v as signs for the spirants. 

IE. j,,i, V.2* 

I. E. y andy disappear in Gadelic, but are preserved in the 

Brittonic as i. Thus loc, heal, 0. Ir. iccaim, W.jach, I. E. yakos, Gr. 

afcos, Skr. j/d^as ; see deigh and dg. For I. E. j, compare G. ebrna, 

.for eo-rna, *jevo-, Gr. (nd, spelt, Skr. ydva; also eud., jealousy,, 

*jantu-, Gr. f^/tos, zeal, Skr. yatnd. 

I. E. V is thus dealt with : — 
(1) Initial v : G., 0. Ir. /, W. gw, as in G. /alt, hair, Iv.folt, W. 
gwalt ; also faidh, Lat. vdtes, feachd, fear, Lat. vir, fiadh, 
fichead, fine, flodh, with succeeding consonant mflath {*vlati-), 
fliuch, fraoch, fras, freumh, etc. 

23 ^ See Supplement to Outlines of Gaelic Etymology. 


(2) Intervocalic v. This disappears in G. leaving the vowels to 

coalesce with varying results, thns : — 

a. -ivo- produces ed, as in bed, *givo-s, Lat. vivus, or la m 

hiadh {*blvoto-n, of. dia), dian. 
h. -evo- produces eb, as in ceo, *skevo-, Eng. shower ; deb, W. 

dywy, *devo-, Lat. fdmus, ebrna. Stokes gives cliu as 

*klevos, Thiirneysen as Moves-. 

c. -ovi- gives nuadh, *novios, -ovo- in cro {*Jcrovos), -ovn- in bg. 

d. -avi- in ogha {*pavios) ; datli {*david) ; -avo- in clb. 

e. -eivi- in gle, -eivo- in dia. 

(3) Post-consonantal v. 

a. After liquids it becomes hh. See garbh, marhh, searbh, 

tarbh, dealbh, secdbh, tneanbh, banbh. 

b. After explosives it disappears save after d, (gv) : feadhbh, 

widow, 0. Ir. fedb, taobh, baobh. For gv, see g below. 

c. After s, it sometimes disappears, sometimes not. Thus 

piuthar is for *sves6r, 0. Ir. siur, whereas in searbh 
(*svervo-s), solus (but follas), seinn, etc., it disappears. 

The Consonant Vowels. 

These are r, I, n, m; f, l, n, in. The regular representation 
•of r, I in G. is ri, li (mvitated forms being rea, rei, lea, lei). See 
the following regular forms : bris, britheamh, fri, lit ; also the 
modified forms — bleath, bleoghainn, breith, cleith, dreach, leamhann, 
leailiaii (?), sreath. 

The numerous Gaelic a forms of I. E. e roots containing 
liquids fall to be noticed here. Some of them Brugmann explains 
as glides before sonants, somewhat thus : G. mair, remain, 0. Ir. 
maraim, would be from mrra-, root mer, Lat. mora ; so sgar from 
sher ; garbh, marhh. 

Add the following : — alt, carbad (Lat. corbis), harr, bard, cairt, 
garg, mall, dall, sgaird (Lat. niuscerda), tart, tar ; fras, flath, 
fraigh, graigh, hraich. With modified vowels in — coille {*caldet-), 
doire, foil, goile, goirid, sgoilt. 

The long vowels f and I appear regularly as ra (1) Id. See Ian 
i*pl-no-, Skr. pumas), slan, tlath, blath. Long p seems to appear 
as ar in dair, maireach, faireag (1).^^ 

Vocalic n and m may be looked for in G. samhaU, which 
Brugmann explains as siiimlli-s, in tana, thin ; reversed in magh 
and nasg. 

Compensatory n plays a great part in G., appearing usually as 
■eu {ao). We have ceud, hundred, W. cant, deud, W. dant, teud, 

-■' See iVupplcmcnt to Outlines of Gaelic Etymology. 


eud, ew], eudann, eiginn, geug. The negative n appears before 
vowels as an, before c, t, and s, as eu, M : eutrom, e'islean, &c. 
The most curious result arises from -ngm-, which ends in G. as 
eu7n-; see ceum, W. cam, hum, W. lam, and add teum, W. tarn, from 

Before the medials 6, d, g, both n and m become in (ion), im 
(iom), and original in retains its i (of. fionn). Thus we have im-, 
iom- from Tnii, Lat. ambi, also \m, ionga, imleag, ciomach. 

I. E. "r" and "I" Liquids. 

Gaelic r and I represent the I. E. liquids r and I. Initially we 
may select rdtmh, reachd, ruadh, run, loch, laigh, lahhair, leth ; 
after p lost — ro, rdih, lamh, Ian, lar. Medially r and I are 
" aspirated," but the sounds have no separate signs — dorus, tulach, ' 
geal, meil, eile, seileach, etc. Post-consonantal r and I appear in 
sruth, srath, etc., cluinn, flinch, slug, etc. In -br, -tr, -dr, the 
combinations become -bhar, -thar, -dhar, while in -cr, -gr, -hi, -tl, 
-dl, -el, -gl the respective explosives disappear with lengthening 
of the preceding vowel. For -si, see below {-U). 

Ante-consonantal r and I preserve the explosives after them — 
ard, hhrd, ceart, neart, dearg, dearc, allt, calltuinn, gilh, halg, cealg, 
olc, etc. 

Gaelic -rr arises from -rs ; see bctrr, earr, carraig ; from the 
meeting of r with r, as in atharrach ; from rth, as in orra from 
ortha, Lat. orationem. Again -II comes from -si, as in uaill, coll, 
ciall, etc. ; especially from -In-, as in follas, hall, feall, etc. ; from 
-Id-, as in call, coille, and many others. 

Gaelic -rr arises from -rp ; corran, searrach (St.) ; Ir. carr, 
spear, cirrim, I cut, forrach, pole. KZ. 35. 

I. E. "n" and " m" Nasals. 

I. E. n and m appear normally in G. as n and m, save that I. E. 
terminal m in neuter nouns, accusative cases, and genitives plural, 
became in Celtic n. (1) Initial n appears in nead, Eng. nest, 
neart, neul, nochd, naked, night, nathair, nuadh, nasg, na, not, etc. 
(2) After an initial mute, n appears in cnaimh, cneadh, end, gnath, 
etc. After s, in snath, sniomh, snuadh, snigh, sneachd. After h it 
changes the b into m {mnatha for *bn6is). (3) Intervocalic n is 
preserved — bean, Ian, maoin, dan, run, dun, scan, etc. (4). Pre- 
consonantal n is dealt with variously : 

a. Before the liquids, n is assimilated to m and I, and dis- 
appears before r. 


b. Before the labials, n becomes m in modern Gaelic. Before 

t, c, the n disappears with lengthening of the previous 
vowel, as in cevd, first, breug, c6ig. Before d and g, 
it is preserved, as in eumhang, fulaing, mumg, seang, 
but it assimilates d—fionn {*vindo-s), borni, inrtr, binn. 
Por -ngm, see under n and g. 

c. Before s, « disappears as before t and c. Compare rnios, 

feiisag, grlos, slos. 

(5) Post-consonantal n disappears after I, leaving II (see under I), 
but is preserved after r, as in cam, ebrna, tighearna, etc. 

a. After s, that is, -sn becomes -nn ; as in dronn for 

*dros-no-, down, uinnsean, cannach, biminne, etc. 

b. The mutes, t, d, c, g, p, disappear with compensatory 

lengthening of the previous vowel : -tn^, as in eun, buan, 
uin ; -dn-, as in bruan, smuain ; -en- is doubtful — cf . ton, 
also sgeun, hrewn, lebn ; -gn, as in feun, bron, uan, srbn ; 
-pn, as in suain, cluain, cuan ; -pn ? tepno = ten ; 
apnio = dne (Lit. aps) ; lipn = len, follow ; but 
supn = suan ; copn = cuan (Stokes); cn,ign, and tn initial 
become r in pronouncing ; but the vowel is nasal — 
gnath is grath with nasal a ; bn becomes mn, as 
in mnaoi, pronounced rnraoi ; even snath becomes 
dialectally srcbth, especially in oblique cases. 

c. After b, that is, bn changes into mh-n, as in domhan 

{*dubno-), sleamhuinn. 

The G. combination -nn arises therefore from (1) m before n, 
(2) n before d, and (3) from -sn ; or (4) it is a doubling of n in an 
unaccented syllable at the end of a word {tighinn, etc.), or, rarely, 
of a one-syllable word like cinn, cluinn, linn. In Islay, -in 
becomes -inn ; duinne is for duine ; minne gen. of min, etc. In 
general, gloinne is comp. of glan. 

Initial m appears in Tn\os, muir, fnil, maide, etc. Before the 
liquids r and I, the m becomes 6, as in braich, brath, brugh, blath, 
bleith, bleoghainn. Intervocalic m is always aspirated — geimheal, 
amhuil, like, cruimh, amh, damh, cnaimh, iamh, caomh. In 
combinations with other consonants, various results occur : — 

(l)'Pre-consonantal m. 

a. Before liquids, ni is preserved in an aspirated form 
{geamh-radh, etc.), but there are no certain ancient cases. 
Of course, m before m results in preserved m (cf. amadan, 
comas, comain). 


h. Before s, m should disappear, but no certain Celtic cases 
seem to occur. In the historic language, m before s 
results in mp or p as usually pronounced, as in rompa 
for rom + so, that is, *rom.-sho ; so iompaidh, umpa. 

c. Before the explosives. Original mh is now m, as in the 
prefix im-, torn-, in imleag, torn. I. E. m before t and k 
(q) became n (as in ceud, breug), and disappeared, with 
compensatory lengthening. Compare also didean, 
eiridinn. Prehistoric mg, md fail us ; in the present 
language both appear aspirated {mhgh, mhdh). 

(2) Post-consonantal m. After the liquids r, I, and ,n, the ni is 
preserved. Whether an intermediate s is in some cases to be 
postulated is a matter of doubt (as in gairm, from * gars- 
men 1 W. garm). See cuirm (W. cwrw), gorm, seirm, deitm, 
calma, ainm, meanmna, anmoch. 

After s, m becomes in the older language mm, now m ; dndm 
comes from ''■dros-men. But s is very usual as an intermediate 
letter between a previous consonant and m : many roots appear 
with an additional s, which may originally have belonged to an -es 
neuter stem. We actually see such a development in a word like 
snaim, which in E. Ir. appears as snaidm (d. snaidmaimm), from a 
Celtic *snades-men. In any case, a word like rvaini postulates a 
Pre-Celtic *roud-s-men. See also gruaim, seaman, r^im, lorn, trom. 

After the explosives the m is aspirated and the explosive 
disappears, as in the case of freumh (vrdmd) ; but seemingly the 
accented prefix ad- preserves the m, : cf. amas, amail, aimsir. 

Preserved G. m, intervocalic , or final, may arise from (1) m or 
n before m, (2) s before m (also -bsm, -tsm, -r/sm, -csm, -gsm), (3) 
-ngm, or -i),gin, as in cev/n^, leum, beuvi, geum, or -ndm as in teum, 
(4) ng becoming mh as in \m, turn, tom, etc., or (5) mb (-mhh), 
as in im-, iom-. 

§ 5. Vowel Gradation oS Ablaut. 

The most characteristic roots of the I. E. languages are at least 
triple-barrelled, so to speak : they show three grades of vowels. 
The root pet, for instance, in Greek appears as pet, pot, pt 
(TreTonat, fly, Trorao/tat, flutter, TTTepov, wing). The first grade — 
e — may be called the " normal " grade, the second the " deflected " 
grade, and the last — pt — the "reduced " or " weak " grade. The 
reason for the reduced grade is evident ; the chief accent is on 
another syllable. Why e interchanges with o is not clear. The 


leading I. E. series of vowel gradations are six in number, a» 
follows : — 

Normal. Deflected. Weak. 

1. e-series e o nil 

but ei oi i 

2. e-series e 6 s 

3. a-series a 6 s 

4. 6-series o o si 

5. a-series a a (a) 

6. o-series o o (o) 

Corresponding to the e, o, nil series are the two " strong " 

vowel grades e, 5, as in sed, sit, sod, sed, sod, si-zd, found in Latin 
sedeo (sed), G. suidhe (sod), G. s\th [properly s\dK\, peace (sed), 
Eng. soot (sod), Lat. sldo (si-zd). 

The e-series in full is as follows : — 

Normal. Deflected. Weak. 

e simple e o nil 

ei ei oi i 

eu eu ou u 

er (or el, en, era) er or r 

To all these correspond "reduced" long forms — to ei belongs %, to 
eu belongs u, and to the consonant- vowels correspond the long f, I, 
n, m. We may also here add the triple ve, vo, u {vet, vot, ut, as in 
G. feitheajnh, iline, uiridh ; vel, vol, ul as in fait, 0. Ir., Mod. Ir. 
folt, olann). 

Some Gaelic examples will now be given. 
(1) The e-series. G. eadh, uidhe from *pedo-, *podio- ; tigh, tugha, 
from *tegos, *togio- ; geas, guidhe from ged, god ; cleachd, 
cleas, cluich, etc. In ei we have the complete set meit, moit,. 
mit in meith, maoth, meata or miosa ; further cliatkach, claon 
from klei, kloi ; fianuis, fios from veid, vid ; gaoth, geamhradh 
from ghoi, ghi ; and others. The diphthongs eu, ou cannot 
be differentiated, but the short form of the root occurs, as in 
ruadh, roduidh from roud, rvdd; huail, buille from hhoud, 
hhud ; cluas, cluinn from kleu, hlu ; nuadh, nodha (?) The 
liquids show the changes also : heir, hreith from ber, br, and 
in the sense of speech we have also hrath, judgment (hftu-). 
The root pel is especially rich in forms : iol (*pelu-), uile 
(*polio-), lion {*pleno-, Lat. plenus, from pie), Ihn (either 
*plono, plo, Eng. flood, or *pl-no-, from pi-), that is, root 
forms pel, pol, pi, pie, plo, pi, meaning " full." In n we have 
teann, tana {*tendo- tnnavo-, according to Brugmann), and 
teud ; from gen we get the long forms gne in gnwmh and gno 
in gnhth. In nem we have neamh, heaven, 0. Ir. nem, and 
ndimhaid, foe, from nom (Gr. vM/iatu). 


(2) The e and other series. One of the best examples of the e 
series is sne, sno (md), spin, which gives smomh (*snemu-) and 
snath, thread {*sn.dtio-). From se comes slol (*sSlo-) and, 
possibly, sath, transfix (sSto-). The a- series is not diifer- 
entiated in G. nor is the o- series ; but from a short we get, 
among others, the root acf, lead, in aghaidh, etc., and dg in 
agh, success, aghaeh, warlike. The diphthong ai has as its 
" reduced " grade i. The name Aodh in Mackay represents 

0. Ir. Aed, aed, fire, Gr. at^co, I burn. 

§ 6. The Spirants. 

The I. E. spirants were j, v, s, and z. We have already dis- 
cussed ;■ and v tinder the heading of semi-vowels, from which it is 
difficult to differentiate the consonantal j and v. Here we deal 
with s and z, and first with s. 

(1) Initial s. Before vowels and the liquids, I. E. s remains intact 

in Gadelic. In Brittonic s before vowels becomes h ; before 

1, n, and m, it disappears, while before r it or its resultant 
effect is preserved (see sruth, srath, srdn). 

a. I. E sv appears in Gadelic as s usually," more rarely as / 
and ^ or « ; in W. the form is chw. See searbh, seal, se, 
sihh, seid, etc. The G. piuthar appears in Ir. as siur, 
fiur, from *svesor, while pill {*svelni-) gives fill and till ; 
compare also s^isd (teis). 

h I. E. sp (spK) is treated in Celtic much as sf. And spr 
appears as sr ; cf. srdn, straighlich, slis, sonn, sealg, sine. 

I. E. St appears in Gadelic as t, as in tigh, tct, tighinn, taois. 
But str, stl, become sr, si, as in srath, sreothart, sreang, 
slios, slat, sioinn, slaid. Some hold that st may appear 
as simple s, which is the case in Welsh, but the instances 
adduced can be otherwise explained (cf. seirc, sAil, 
searrach (St.), seall). 

I. E. sq, sqh, appear in Gaelic as sg, 0. Ir. sc, as in sgath, 
sgath, sguir, etc. The W. precedes the sg with a j/ as in 
ysgwyd, Ir. sgiath, G. sgiath, shield : I. E. sqv is in W. 
chw, as G. sgeul, W. chwedl, sgeith, W. chwydu. 

I. E. skn appears in Gaelic as sn, as in sneadh. 

(2) Intervocalic s. This becomes h and disappears ; compare tagh 

{*to-gusd), do-, chl, etc. 

(3) Terminal s disappears altogether ; but in closely connected 

combinations of words its former existence is known from the 
so-called euphonic h, as in the article genitive feminine and 


nom. plural before vowels {na h-oighean = *sen'1di augeis), also 
0' H- of Irish ; and it may be the origin in most cases of 
prothetic s. 

(4) Pre-consonantal s. A prehistoric case of -'■t is not forth- 

coming, but eirich comes from *ek-s-regd. Before I, m, and n 
the s disappears, and the liquid is doubled (m of Gaelic being 
for older mm), as already shown under these letters. Medial 
sv appears as / in the older language (see seinn), and it is 
still seen in tibhann (*to-sven-), feahhas. 
Before the explosives, s is preserved before the tenues, which 
in the modern language become mediae. The combination 
sp is not certain ; but -sc becomes -sg (see fasgadh, seasg, 
measg, etc.), st becomes s (older ss) simply, as in seas 
( = *sisto-), fois, fan, dos, etc. Before the medials s becomes 
z, which see for results in Gaelic ; *sg becomes g ; sp becomes s. 

(5) Post-consonantal s. After the liquid r the s is assimilated to 

the r, and the result is rr, as in barr, earr, etc. From -Is- 
seemingly s results, at least in the later language ; -ms, -ns 
become s with compensatory lengthening for the previous 
vowel ; -rfs" becomes t, as in an t-each ( = *si'ndos eqos) ; Thn. 
adds fitir { = *vid-sar). For m-sh = mp, see u.nder m. 
The explosives combine with the s and disappear into 0. Ir. ss, 
now s, as in tiasal ( = *oups- or *oui:s-y, his, leas l*led-so-), 
lios, as, out ( — eks), and many others. 

Gaelic preserved s intervocalic, therefore, arises from (1^ st, as- 
in seas ; (2) from -ms, -ns, as in mhos ; and (3) from -ps, -ts, -cs. 
Gaelic -st arises from this s by a sort of modern restoration of 
previous st, only, however, x may also become modem st (as in 
aiste, now aisde, out of her). Final x disappears, as in caora, se. 

I. E. 2. 

Even in I. E. this is assured only before the medial explosives. 
Thus G. nead, nest, is from I. E. nizdo-s : so maide, brod, cead, 
gad, seid. Again -zg seems to have developed in G. into g ; compare 
beag, biog, meag, griogag, eagal ( = ex-gal-), rag. 

§ 7. The Explosives or Mutes. 

The I. E. explosives formed a possible sixteen in number 
between tenues, medise and the double set of aspirates {j>h, bh, th, 
dh, kh, gh, qh, gh). The tenues aspirate were "rare and of no 
importance" in the resulting languages, save only in Sanskrit and 
Greek. The medise aspirates are the predecessors of aspirates of 
the modern languages. But in the Celtic languages these medise 


aspirates were merged into the medise themselves, so that h and 
hh appear in Celtic as b, d and dh as d, g and gh as g, and g and 
gh as g. The Balto-Slavonic, in this matter, shares the peculiarity 
of the Celtic. 

All the explosives, when intervocalic, are " aspirated" in Gaelic 
— p to jph, b to bh { = v), t to th ( = h), d to dh ( = y), e to ch, g to 
gh, { = y); the corresponding Welsh changes are the tenues to 
mfidiae, and the medise to /, dd, and nil in the case of g. Inter- 
vocalic preserved explosives in Gaelic arise from a doubling of the 
explosive, the cause of which in many cases is obscure. The fol- 
lowing are the leading cases and causes of intervocalic G. mutes : 

(1) Doubling of the explosive in the course of inflection or word- 


a. Inflection. The participle passive in -te preserves the t or 
d of the root as t; thus \caith gives caithte,^ bath (for badh) 
gives bdiite, rkdh gives raite, etc. 

h. Word-building. The prepositional prefixes which end or 
ended in a consonant preserve the succeeding explosive ; 
even vowel-ending prepositions like air {*are), aith- 
(*ati) do the same, if the accent is on the preposition. 
Thus — abair is for ad-ber, aitreabh is for ad-treb, aidich 
is for ad-dnm, faio for ad-ces-, agair for ad-gar. In the 
way of affixes, we have ruiteach from rud-t and ruicean 
from rud-c, creid from *cred-dh6 ; compare the compounds 
boicionn, laoicionn, and craicionn. 

(2) After sunk n or m. Thus deud comes from dnt, and so with 

ceud, tend ; ceud, first, from *cento-, so send ; eug from tjJco-, 

(3) After sunk spirant z. This is assured for zd, as in brod 

(*broz-do-, ^iorse broddr), cead, gad, maide, nead ; but zg 
giving g is doubtful — eagal seems for *es-gal or * ex-gal-, 
beag for gvezgo-s ( Lat. veseus), meag for mezgo-. 

(4) Cases corresponding to double explosives in other languages : 

cat and Lat. catta (borrowing ?), cae and Gr. ko-kkt]. Compare 
also dug. 

(5) Doubtful cases. Many of these cases can be satisfactorily 

explained as due to suffixes immediately affixed to consonant- 
ending roots. Thus brat may be for brat-to-, trod for trud-do-, 
toe for *yak-ho-, breac for mrg-ko-. Even suffixes in -bho- and 
-go- (Eng. k in walk) are not unknown, and they might 
account for reub {*reib-bo-, *reib-bho-, Eng. reap, rip), slug for 
slug-go-, etc. Dr Whitley Stokes has given a different theory 
founded on the analogy of a Teutonic phonetical law, stated 


thus by Brugmann : " hn, dn, gn became hb, dd, gg before the 
principal accent in primitive Teutonic, thence pp, it, kh (by 
Grimm's law), which were further treated just the same as 
pp, it, kk, which had arisen from pn, tn, qn, and from I. E. 
hhn, dhn, ghn, gh'n. . . . 0. H. G. shiccho, slukko, 
glutton \^8luk^no-\ M. H. G. sluchen, gulp, have hiccup, 
allied to Gr. )^v(<ji, X.vyya,vdofj,ai, I have hiccup." These last 
words are allied to G. slug, which Dr Stokes refers to a 
pre-Celtic *slug-nd-, the accent being on the suffix -no-. The 
weakness of this hypothesis lies in the fact that uniform 
results are not found from it. Thus breac, from mrg-no-, 
should be hreag, not hreac, on the analogy of dug. 

I. E. p. 

Initial and intervocalic I. E. jo disappears in Gaelic, as in 
athair, Lat. pater, eun for *pet-no-, eadh for pedo-, iasg against Lat. 
piscis, ibh against hiho (for pibo), Ian against Lat. plenus, lar and 
Eng. floor, etc. For intervocalic p, see fo (*upo), for, teth, 
caora, (^kaperax), saor, (*sapiros), etc. 

Lat. and G. agree in the initial of the numeral iive — quinque 
and ctig, though the L E. was penqe. In feasgar the G. guttural- 
ises an original vesperos without Latin countenancing it. Initial 
sp appears as s ; see sealg, spleen, sonn, sliseag, sine, sir. 

When p appears before the liquids and t, c, or s, it is not lost 
in G. ; it leaves its influence either in a new combination or in 
compensatory lengthening. Thus suain is for supno-s, and see 
cluain, cuan. G. dias seems from *steip-s-d, W. twys, and vasal 
may have had an original form like v\j/r)\6'i, Eng. up. (Of. teanga 
and dingua). In seachd, Lat. septem, the p is gutturalised ; we 
may add here *neachd, 0. Ir. necht, Lat. neptis, Eng. niece ; 
crev^hd, dreachd. Possibly leac may be for lep-kd, 

G. intervocalic p is, of course, due to some combination. In 
leapa, genitive of leabaidh, it arises from *leb-tha ; and we must 
explain similarly tap (*tabaidh arising from *tab-tha) ; so raip, 

For t taking the place of p through an initial h compare the 
derivations offered for tore, turlach, tuil, tlcim, tliis for liiths. 

I. E. 6, bh. 
These two become b in Gaelic and the other Celtic languages, 
I. E. b is rare in any language ; in G. it appears in ibhim {*pib6). 
treabh, domhain and dritchd {*dhreub-tiir). 

(1) Initial I. E. bh, G. b. See beir, balg, ball, ban, blath, bloom, 
bragh, biuthainn, buaidh. 


(2) Intervocalic I. E. hh, G. bh { = v), 0. Ir. b, W.f. 

crdbhach, dubh, gubhal. 
(3). Pre-consonantal bh or 6. 

a. Before r it remains — abhra, gabhar, dobhar, Gaul, dubrum. 

b. Before I it disappears with compensatory lengthening — 

neul for neblo-s. 

c. Before n it becomes mh now — sleamhuinn is for *slibno-s, 

Eng. slipper^/ ; so domhain. These are I. E. b. 

d. Before t, I. E. b becomes ch as in driichd. 

{4) Post-consonantal b, bh. It is preserved after the liquids r and 
I — carbad, cearb, earb, gilb, sgolb. After m, it preserves the 
m, as in im-, iom- from rnbi, ambi. After s it is preserved 
in eabar ; after d in abair, leob, faob, aohrann ; perhaps after 
g in leabaidh, *leg-buti- if). 

{5) Gaelic intervocalic b. In reub and gob we seem to have a 
suffix -ho-, *reib-bo-, gob-bo ; also cliob from clib-bo-, root qlg, 
Gr. KoXo/Bo^, stumpy (?). Oftenest b is produced from a 
previous d, especially of the prefixes — as abair, abadh, faob, 
etc. (see the paragraph above). 

I. E. t. 

Initially this is Celtic t ; intervocalic, it is aspirated, and 

otherwise it is variously modified. 

(1) Initial t, G., 0. Ir., W. t. See, among many, tiugh, tar, teth, 
teich, tais, tor a, tlMh, tnuth, tri, treabh. 

{2) Intervocalic t, G. th { = h), 0. Ir. th (d), W. d. See athair, 
mathair, ith, roth, ceithir, leth, etc. Sometimes in non- 
accented syllables it appears as dh, as in biadh from *bivoto-s, 
and this is always the case with the infinitives in -atu- 
(glan-adh). Irregularly faidh for faith. 

(3) Pre-consonantal * not initial. Before r it is preserved, as in 

criathar, briathar, etc. Before I it disappears with com- 
pensatory lengthening — sgeul, W. chuiedl, bl, beul, etc. ; so 
before n, as in eun. Before s the t disappears and the s is 
preserved, as in miosa, ris, sd,s. Words like fios are from 
vid-s-tu-, formerly explained as from vid-tvr. Before another 
t, t is preserved in the resultant < of G., as in ite, etc. ; -td- 
seems to become -dd- ; -tc- becomes 0. Ir. cc, G. c, as in 
freiceadan ; -tg- becomes gg, that is g, as in freagair. 

(4) Post^consonantal t. After r and I it is preserved, as in heart, 

ceart, ceirtle, alt, fait ; after n and m it sinks to d, as in ceud, 
etc. As seen, -bt becomes -chd,, as in druchd, while -pt is in 
seachd. After c or g, the t sinks in G. to d. preserving the 


guttural as an aspirate : oclid, nochd, boclid, reachd. 0. Ir. 
has -cht here and W. th. 
(5) Gaelic intervocalic t. The * of a root is preserved when the 
suffix begins in t, as [in caithte, spent,] in ite, Ir. ette, *pet-tid, 
lit, *plt-tion-. The d of the affixes preserves it, as in aitreahh, 
taitinn, ruiteach, riit. The ( of the following does not 
belong to the ultimate root : ciotach, *sqvi-tto-, Eng. skew, 
croit, root kur, lot, root lu. 

I. E. d. dh. 
This is a uniform Celtic d initial ; Gaelic dh between vowels 
and W. dd. 

(1) Initial d, dh. See deas, dearc, deich, druim, dim, damh, etc., 

for d ; for dh, dubh, domhan, dearg, dorus, dall ; also dlighe. 

(2) Intervocalic d, dh. See fiodh, *vidv^, eadh, suidhe, fiadh, 

guidhe, etc. 

(3) Pre-consonantal d, dh non-initial. Before r, I, n, the d dis- 

appears with compensatory lengthening, as in aireamh 
{*ad-Hm-) &roK, arach, buail, {*boud-lo-), but buille is for 
*bud-s-lio- ; amu^in for smoud-no-. Before' m it sometimes 
disappears, as in freumk, *vrd-m,d, but with an accented 
prefix the d and m become m, as in aimsir, amal, amas. With 
s it coalesces into s, as in musach, or in uisge for *ud-s-qio-, or 
flos for *vid-s-tu-. Before the explosives, with b it coalesces 
to bb, now b, as in abair, etc. So with t, as in aitreabh ; with 
d, as in aidich ; with c, as in faie ; with g, as in agair. 

(4) Post-consonantal d, dh. The liquid r preserves a following d, 

as in ard, bard, sgdird, brd, etc. It assimilates with I, as in 
coille, call, moll, mullach ; and with n, in fionn, 0. Ir. find, 
bonn, 0. Ir. bond, binn. For zd, see next paragraph. The 
explosives before d are unusual, save t and d, for which see 
next paragraph. 

(5) Intervocalic G. d. There are three sources at least for this 


a. The d from nt in ceud, teud, beud, etc. 

b. The d arising from the spirant z before d, as in brod, 

*brozdo-, cead, gad, maide, nead, druid. 

c. From -dd- as in creid, goid, rodaidh, trod, etc. ; also aidich, 

* ad-dam,-. 

I, E. " h " and " q." 
These appear in G. uniformly, as c, ; but in the Brittonic 
languages q, if labialised, becomes jp as in Greek. 
(1). Initial k. See cluinn, cit, ceud, hundred, cac, cridhe, caomh, 


Initial q simple. See caraid, W. cdr, ceud, first, W. cynt, coille, 
W. celli, cas, W. coes, coileach, W. ceiliog, etc. 

Initial q labialised, that is, qv : casd, W. pas, ciall, W. pwyll, 
ceithir, W. pedwar, ceafin, W. pen, coire, W. pair, co, W. ^ja, 
cruimh, W. jo?'y/. 

It seems clear that G. gr at times represents I. E. h, q, as W. 
has the latter. Compare G. geug with W. cainc, Skr. ganku ; 
but W. ysgainc shows the reason for the anomaly — an s 
initial has been dropped, and in dropping it the G. reduced 
c to g. Further compare garmainn, giomach. Cf. dias. 

(2) Intervocalic k, q. The G. is eh, W. g, b. Compare cruach, 

W. crAg, fichead, deich, loch ; also each, W. ebol, seach, 
W. heb, etc. 

(3) Pre-consonantal Ir, q. Before r, I, n, the c disappears with com- 

pensatory lengthening as in rfewr, Lat. dncrima, meur, dtial, 
muineal, ton ; and compare Prof. Strachan's derivations for 
meanan, brewn, chin, leana. With s, the result in G. is s, 
0. Ir. Si, W. ch, as in uasal, W. iichel. Before explosives, 
cb, cd, eg do not appear ; ct becomes chd, for which see 
under t (4) ; for c-c, see paragraph (.5) here. 

(4) Post-consonantal k, q. After r and I, the guttural appears as 

c, as in cearc, uireean, male, olc, falc, etc. After n (m), it 
sinks to g, with a preceding long vowel, as in eug, breug, 
already discussed. Aftei» s, the c is preserved, but in G. it 
is written as g, as in measg, nasg, teasg, etc. After 
explosives, the t and c^ of the prefix or root preserves the c 
following, for which see under t and d pre-consonantal. 
For c 01 g before c, see next paragraph. 

(5) Intervocalic Gaelic c. It may arise from -tk, -dk, -kk, -gk. 

From -ik in freiceadan {\frithrCom-dt-dn) ; -dk in faie, 
aearach, 'ruicean, aeuinn ; -kk in mac, *'m.ukkus, cac, 
craicionn, loc, leacainn ; from -gk in bac, boe, breac, enoc, 
gleac. The word mac, son, postulates a Gadelic makko-s as 
against the Ogmic maqvi (gen.) and W. mab ; it is difficult 
to account for the G. form. 

I. E. g, gh ; g, gh. 

These consonants all, save in one case, appear in G. as g, 
aspirated to gh, and W. shows g and nil in similar circumstances. 
The exception is in the case of g, which when labialised, becomes 
G. and W. b. But gh, whether labialised or not, becomes g in G. 
(1) Initial I. E. g : in guth, gin, gnhth, geim.heal, gb. I. E. gh is 
in geamhradh, gabh, ghg, geal, white, I. E. g simple appears 
in geal, leech, goir, goile, gearan, guala, gradh ; I. E. gh in 


gar, grian, gaol, guidhe, geas, guin. Labialised g appears in 
hean, Eng. qiieen, bior, beb, b6, bra, quern, braghad. 

(2) Intervocalic Celtic g. See deigh, aghaidh, greigh, truagh, 
Ueoghainn, tigh, bragh, etc. In the termination of words it 
appears often as ch: teach, (*tegos), mack, {*magos), imlich, 
im[th]ich, dirich, fuirich. Intervocalic g labialised does not 
seem to exist in modern G. 

{3) Pre-consonantal Celtic g. Here -gr, -gl, -gn, become -r, -I, -n 
with vocalic lengthening, as feur, *vegro-, ar, nclir, fuar, al, 
fual, feun, *vegno-, srbn, uan, tain, bron, etc. Before m, g 
is found in the combination ng^m, which results in m with a 
preceding long vowel, as in ceum, leum, geum. Before s it 
becomes x and modern s, W. ch, as in vasal, W. uchel, as for 
ex, OS, deer, W. ych, cas, las, uiseag. Before explosives the g 
is variously preserved : -gb, -gd may be passed over ; -ct, -gt 
appear as chd, as in seachd, bliochd, smachd, nochd, sneachd, 
etc. ; -gh ends in -kk, now c, for which see pos1>consonantal k ; 
-gg appears as g, as in slug, bog, clag, lag, slige, smugaid. 

{i) Post-consonantal Celtic g. After r and I the g is preserved 
in G., but often in W. becomes y ; see dearg, fearg, searg, 
garg, lorg, balg, cealg, dealg, tulg. After n ordinary g is pre- 
served, as in cumhang, long, muing, seang, fulaing. But 
labialised g became b, and then coalesced with the n into mm, 
now m as in \m, butter, " Lat. unguenium, turn, cam, torn, 
ciomach, and in modern times cum, keep, from *congv in 
congbhail. For ng-m see the foregoing paragraph. For tg 
see the next paragraph. After the explosives, the g is pre- 
served in the combinations -tg [freagair), -dg (agair), and -gg, 
which see below. 

{5) Intervocalic Gaelic g. It arises from -sg firstly, which in pre- 
Celtic times was -zg, as in beag, mogul, griogag, meag, eagal, 
etc., which see under I. E. z above. From the explosive 
combinations we have tg in freagfiir, */rith-gar-, eagna, eagar ; 
dg in agair, agus. The -gg must arise fromi a suffix in -go-, 
which was operative in early Gadelic, if we discard Dr Stokes' 
view already set forth. Cf. Eng. walk, hark, lurk, skulk, 
smirk. For this -gg see paragraph third above. 
Intervocalic g may arise from a lost n before c, as in breug, gtug, 
eug, etc. The previous vowel is lengthened save in a few 
cases where the word — or sentence — accent has brought about 
a short syllable. Thus thig has short i, and in G. leig is 
short. This is regularly the case with the results from the 
prefix con, confused with cos, as in cogais, 0. Ir. concubus, 
cadal, cagar, eogadh, etc. 

outlines of gaelic etymology. xxix. 

§ 8. Accent. 

In Gaelic, only the stress accent exists, and it is placed always 
on the first syllable. The accent of the Old Gaelic was likewise 
on the first syllable, save in the case of the verb. Here in the 
compounded verbs the stress accent rested on, as a rule, the 
second syllable ; but the imperative placed the accent on the 
first syllable, and this also took place after the negative and inter- 
rogative particles and after the conjunctions gvln and na'n (da'n). 
Thus /ai'c, see thou, is ior f-aid-c, with accent on the preposition 
ad, for it is imperative ; the future chl stands for the old present 
at-chi, videt, where the accent is on the root ci. Again in cha'n 
fhaca the negative brings the accent on the prefix ad, that is, 
f-ad-ca. When the accent is on the prefix, its ending consonant 
and the initial consonant of the root coalesce and result in a pre- 
served G. intervocalic consonant, but the root suffers truncation : 
when the accent is on the root, these consonants are aspirated, and 
the root is preserved. The ten irregular verbs in G. present 
sufficient illustrations of this rule. The preposition con, when 
accented, was always con, when unaccented it was com (comh). In 
the unaccented syllables, long vowels become short [aireamh from 
*dd-rim, anail for 0. Ir. andL), and in many cases change com- 
pletely their grade, as from small to broad (e.g. cbmhnadh, 0. Ir. 
congnam, from gniomh, and the compounds in -radh and -lach). 


Word-building consists of two parts — composition and deriva- 
tion. The first deals with the compounding of separate words ; 
the second deals with the suffixes (and prefixes) that make up the 
stem of a word from its root. 

(1) The compound may be two stems welded together: righ-theach, 
palace, *rigo-tegos, "king's hoiise"; righ-fhaidh, royal prophet 
— " king who is a prophet" ; ceann-fhionn, white-headed, 
penno-vindo-s ; ceithir-chasach, four-footed ; dubh-ghlas, dark- 
blue ; crannchur, lot, "casting the lot." These are the six 
leading relationships brought out in compounds. In Celtic 
the first stem is nearly always in o-, as Teuto-bodiaci, G. sean- 
mhathair (but Catu-slogi, Mori-dUnum, G. Muirgheal). Con- 
sider the following compounds : iodhlann, mlolchu, dircheard, 
htuirach, ceardach, clogad, bclthach, eilthire, gnath-fhocal, 
moirear, leth-chas, leth-trom, etc. 
The following are common prefixes : ath-, re-, ath-ghlac, re- 
capture ; ban-, she, ba7i-altrum, hantrach ; bith-, ever-, bith- 
bheb, bith-bhuan ; il-, iol-, many ; ion-, fit ; sir-, slor-, ever-, 
fir-, fior-, very, saobh-, pseudo-. 


The following suffixes belong to this branch of word-building : — 
-lach, from *slougo-, now sluagh ; seen iii teaghlach, dbrlach, 

bglach, youth, etc. 
-radh, from *reda, W. rwyd (see riidh) ; seen in reabhradh, 

madraidh, dogs, bigridh, youth, macraidh, sons, rlghre, 

kings, gniomharra, deeds. 
-mhor, -or, from mor, great; it makes adjectives from nouns, 

etc. : llonmhor, etc. 
-ail, like ; from samhail, amhail : rioghail for riogh-amliail, 

-an, diminutive masculine, 0. Ir. dn, Ogmic -agnos, for *apo- 

gno-s, root gen, bear (Stokes) : as in fearan, truaghan, 

-ag, diminutive fem. in G., 0. Ir. -6c (masc. and fern.), from 

be, bg, young : seen in caileag, etc. 
-seach. This feminine termination has been explained by 

Stokes as from 0. Ir. es, a fem. form, with the adjectival 

addition *iqd, and this es he deduces from W. es, which 

comes from Lat. issa. Cf. haiseach, cldirseach, bonnsaeh, 

c^irseach or ciarseach (Ir.). 

(2) The compound may be one noun governing another in the 

genitive : mac-leisg, and all the personal names in mac, gille, 

(3) Uninflected prefixes : 

a. Negative prefixes — I. E. n, G. an before vowels, aineol, 

ion-, in- before h, d, g (iongantas), eu- (ao-) before t, c, s 
(aotrom for d-tr07n, *n-trom,mo-s). 
To this negative add also mi-, neo-, as- (eas-), di- (der- = 

b. Prefixes of quality : do {do-char), and so- (so-cliar) ; and 

the intensive ro-. 

(4) Old adverbial forms and all prepositions. These prepositions 

are often combined with one or two other prepositions. 

ad-, Lat. ad : faic —f-ad-ci ; aireamh ( = ad-rim-). 

aith-, ad-, *ati-, re-, continually confused with the above 

prep, (aith gives accented e as in ^piur ; ad gives a as in 

aca) : ahair {*ad-her-), agair, aithreachas {*ati-rec-), etc. 

Compounded with to- in tagair, tapaidh, taitinn, taitheasg, 

taisg, etc. ; with /o- in fag (fo-ad-gab). 
air, I by, on : air-leag, eir-idinn, bir-dheirc, oir-thir, urchair, 

iirlar. Compounded with com in comhairle ; with to- in 

tairis, tairg, tearainn ; with di- in dearmad , with imm,- 

in inmar-hhaigh, iomarchur. 


as, out, es- : as-eirigh, as-creideamh, eas-hhuidh, ei-rich. Com- 
pounded with air : uireafhhuidh ; with to-, teasairg ; 
with to-for- in tiiairisgevi ; with to-fo-ar in tuarasdal ; 

with to-fo- in tvnsgail. 
eadar, between ; eadar-sgaradh. 
iar, after ; in *iarfaighim, now febraich ; iarogha. 
in, in ; with to- in tional and comh-thional. With a double 

nn in ionnsuidh. 
inn-, ionn-, to, Gaul, ande- : in fionnoglia ; with to- in tionn- 

sgainn, tionndadh (Zeuss). Confused with in, ind, above. 
im-, iom-, about : iomair, iomradh, imich, iompaidh {*imb-sh). 

Compounded with com in caochladh ; with to- in timchioll, 

tiomsach, tiomnadh. 
od-, ud-, out, Eng. out : obann, ohaidh. Compounded with 

aith- in lobairt ; with di- in diiisg ; with fo in fbgair j 

with to- in tohar, tog. 
con-, comh-, co- : coimhead, comaidh, caisg, cogadh. Com- 

poimded with im- in iomchorc ; with com in cogais (0. Ir. 

concubus) ; with to-aith- in teagasg, teaganih. 
di-, de, de : dimeas, dioghail, dlomhain, dvreack ; also deach, 

do-, to : this is the unaccented form of <o-. 
fo, under : in foghnadh, foghlum, falach, fulaing. Com- 
pounded with to- in tdrachd, tuideadh {to-fo-ess-) tuarasdal 

(to-fo-ar-as-), tuasgail (to-fo-as-). 
for, far, super : in forail, forradh, fardorios, farmad, furtachd. 

Compounded with to in tormach, tuaiiisgeul. 
fri-, ri, to, *vrt, Lat. versus ; it appears as frith, fris : in 

freagair, fritheil, freiceadan {frith-coni-). 
ro-, before : in robhas, rosg, rabhadh, radharc. Compounded 

in rug {ro-ud-). 
tar, across, tairm- : in teirig, toirniisg. 

Stem Suffixes. 

The following are the most important suffixes used in Gaelic 
for stem formation : — 

1. 0-, d-, as in cid {*ciUo-), aitreabh, cas (*coxd). 

2. tro-, tlo-, trd-, tld- : criathar, hrei-tro-, anail, (^ana-tld), sgeul, 


3. jo-,jd-, ijo-, ijd-: eile, suidhe, {*sod-i-on). See no-, ro-, tjo-, sqio-. 

4. V0-, vdr, UVO-, uvd- : tarbh (*tar-vo-), each (*ek-vo-), bed, (bi-vo-). 

5. no-, nd-, nno-, eno-, ono- : Ian, sldn, duan, domhan, leathan 

(letano-s). It is secondary in iarunn ; cf. tighearna (*teg- 


6. mo-, md- : trom, lorn, caomh. 

7. ro-, rd, rro-, etc. : slor, m6r, lar, ar, hodhar. Here comes the 

Gaelic numeral stem -ArcHi, as aonar, one person, c6ignear, 
five persons ; it is allied to Lat. -drius, -drlutn, Gaelic -air, 
-eir, denoting agents or doers — clarsair, harper, etc. 

8. tero-, ero- ; in sinnsear, uachdar, eadar. 

9. lo-, Id-, llo-, etc. : coll (*cos-lo-), siol, neul, ciall, giall. 

10. dhro-, dro-, dhlo-, dlo- : odhar, uallaeh. 

11. hho-, hhd-: earh, goh {*goh-bo-). 

12. to-, tdr. This is the participial termination in most I. E. 

languages. In G. it is used for the past passive. Also in 
the adjectives nochd, hochd, gnath, etc. ; nouns dligheadh, 
dearmad, gort. 

13. tjo-, tjdr : Gr. d/xySpocrios. This forms the passive participle in 

G. : hriste, caithte, etc. 

14. td- of abstract nouns : lobart, now lobairt. 

15. to- comparative. This appears in the ordinal numerals: 

deicheamh, 0. Ir. dechmad, for *dekmmeto-. 

16. ho-, hd-: og, jonng, juvn-ko-. 

1 7. qo-, qd-, qio-, dqo- ; shileach for *suli-qo-s ; cuimhneach, 

creidmheach. Especially the adjectives and nouns in aeh, 
as marcach, huadhach. Further, the form iche {-iqio-s} 
denoting agent ; maraiche, etc. 

1 8. sqo-, sqio- : as in measg, seasg, uisge. 

19. go-, gd : see muing, Danish, manke ; cf. Eng. walk, hark, etc. 

20. Stems in i- : dbird, muir, maith, deigh. In ni-, tctin, cluain, 

buain ; in mi-, cruimh, cndimh ; in li-, samhail, duil ; in ti-, 
fdith, fdith, breitk, bleith, etc. — a form in which some 
infinitives appear. 

21. tdti-, that is, Celtic tat-, tus : beatha, life, *bitils, g. *bi-tdt-os. 

22. Stems in u- : tiugh, jliuch, dttb, loch. In nvr, linn, 0. Ir. lin, 

Unvr ; in <«- there are many — bith, iodh-, fios {*vid-s-tnir), 
guth, cruth ; especially reachd and its like in chd. Here 
come the infinitives in adh {-dtvr). 
In G. -eas, as of abstract nouns, the form arises from tu- being 
added to an -es stem : aois, *aiv-es-tu- ; so dorus, follus. 

23. Stems in -n : cu, ara, \m, ionga. In -ien, there is 'Eire, 

' Eireann. The stems in ti6 are very common ; the oblique 
cases are in -tin- \ see eiridinn, faotainn, etc. : common in 
infinitives. Similarly common is -men, -mon, in ainrn, 
cuirm, druim, leum ; and masculine in britheamh, ollamh, 

24. Stems in -y ; , only the family names athair, mathair, etc. 

25. Stems in -t, -nt : nochd, night : caraid, friend — a participial 



26. Stems in k or q : G. nathair, g. nathrach, so Ihir, lasair, 

cathair, etc. 

27. Neuter stems in -es : teach, leth, magh, gleann. 

28. Comparative stems in -jes, -is-, j6s : mh, greater *md-j6s, sine, 

Skr. san-yas-. 

Adair in tughadair, dialladair, figheadair, breabadair, etc. (?) 

Two or three stems peculiar to Gaelic may be mentioned. 
Adjectives in -idh, 0. Ir. -de, as diadhaidh, come from an original 
-dio-. Endings like maireann, firionn have been correlated with 
the Lat. gerund, itself a much disputed form. The preserved d in 
words \i\eflichead, moisture, 0. Ir. fliuchaidatu, has been variously 
referred to *-antvj- or -ato-tUt ; possibly the latter is its origin. 


A. Declension. 
1. 0- stems. Masc. o-stem bcdl, member. 


Sing. Nom. ball 

Gen. buill 

Dat. ball 

Ace. ball 

Voc. bhuill 

Dual N., A. dk bhg,ll 

G. da bhuill (?) 

D. da bhall 

Plur. Nom. buill 

G. ball 

D. ballaibh 

A. buill 

V. bhalla 

Old Irish, 
ball n- 
da ball 

dib mballaib 


ball n- 







ballu (balls j. Jub.) 




balli (bal-loi) 
ballos (ball )ns) 

Neuter io-stem cridhe, heart. 

S. N., 

PI. N. 









cride n- 



oride n- 


cride n- 


chridheachan chride 






krid on 





2. ^sterns : all feminine, cas, a 



Old Irish. 


S. Norn. 











coxi (coxai) 



coiss n- 






Dual A, 

da chois 

di choiss 



da chois 

da choss 



da chois 

dib oossaib 


PI. N. 






coss n- 














3. i 

-stems. Feminine noun sidl 

, eye. 

S. Nom. 







sulos (siilous) 







Sliil n- 






Dual N. 

da shiil 

di shiiil 



da shuil 

d& siila 



d^ shiiil 

dib sulib 


PI. N. 



suleis (sulejes) 



siile n- 














4. It-stems. Masculine noun bith. 

, world. 

S. Nom. 














bith n- 









bitois, (bitoves 



bithe n- 

bition, (bitovoi 














5. Consonantal Stems. 

(a). Stem in r 

; athair, father. 

Gaelic. Old Irish. 

















athir n- 






Dual N., A, 

da athair 

d& athir 



da athair 

da athar 



da athair 

dib n-athrib 



athraichean athir 



athraicheaii athre n- 







athraichean athrea 

ateras {alerii^) 


athraichean athrea 



Stem in men 

/ neut. ainm, 



N., A. 


ainm ,n- 




anma, anme 







N., A. 

ainmeannan anmann 



ainmeannan anmann n- 

anmenon ' 


aiumeannan anmannaib 


(e). Stem in guttural c , fem. nathair, 

S. Nom. nathair 

G. nathrach 

D. nathair 

A. nathair 

Dual N., A. da nathair 

G. da nathair 

D. da nathair 

PI. N. nathraichean 

G. nathraichean 

D. nathraichean 

A. nathraichean 

V« nathraichean 

nathraig n- 
di nathraig 
da nathrach • 
dib nathrachaib 
nathrach n- 


natracen (natrcn) 

(d). Neuter stem in -es ; tigh, house. 

S. N#, A. tigh teg, tech tegos 

G. tighe tige tegesos 

D. 'tigh tig tegesi 

Dual N. da thigh da thech tegese 




Old Irish. 



da thigh 

da thige 



da thigh 

dib tigib 


PI. N. 






tige n- 






6. Adjectives. 

Adjectives belonged (1) to the o- and the a- declensions, as 
*marvos, *marva, *marvon, now marhh, declined like the nouns 
of 0- and a- declensions ; (2) i- declension, as maith, ^matis, 
*matis, *mati, the neuter nom. being the stem ; (3) u- declension, 
as *tigus, *tigus (?), *tigu, now tiugh ; and (4) consonantal adj., 
*tepens, fe, teit, etc. Comparison was in two ways — (1) caomh : 
0. Ir. c6em, coemiu, coemem : *koimos, *koimj6s, *koimimos ; 
(2) luath : 0. Ir. Math, Itiathither, Mathem : *loutos, *loutiteros, 

The numerals may be seen in the Dictionary in their Celtic 
form : *oinos, *dva, *treis, etc. 

The pronouns are so phonetically gone astray that they cannot 
be restored. 

B. Conjugation. 

Active Voice. Indicative — Present. Verb heir, bear. 
S. 1. beiridh mi berimm berommi* 


beiridh tu 




beiridh e 







P. 1. 

beiridh sinn 




beiridh sibh 




beiridh iad 


berenti (beronti) 





Dependent Present. 

S. 1. 

bheir mi 




bheir tu 




bheir e 



P. 1. 

bheir sinn 




bheir sibh 




bheir iad 


beront' ,, 

first sing, is 

from theme-vowel-less verbs : 

*ber-mi. Cf. orm, tharam 

even ogam, asam. 



Secondary Present or Subjunctive. 


Old Irish. 





no berinn 

berin (?) 



no bertha 



bheireadh e 

no bered 





no bermmis 

berimmiss (?) 


bheireadh sibh no berthe 



bheireadh iad no bertis 

berintiss (?) 



do ghabh 

ro gabus 




ro gabis 




ro gab 





ro gabsam 




ro gabsid 




ro gabsat 










( beir 
1 berthe 



beireadh e 













beireadh iad 



Passive. Indicative — Present. 
S. 3. beirear e berir beretor 

P. 3. beirear iad bertir berentor 

Secondary Present or Siibjunctive. 
S. 3. bheirteadh e no berthe ■ — 

P. 3. bheirteadh iad no bertis — 

Past Tense. 
S. 3. chanadh e ro ch^t 
P. 3. fchanadh iad ro cheta 


S. 3. beirear e berar 

P. 3. beirear iad bertar 

cainte c^te 

cantos, "cantus" 
cantas (n.f.) 




1. cf. Grierson's Linguistic Survey of India as to how far the 
statement is to be limited as embracing India. Concerning Asia 
the statement is to be restricted to living Aryan languages. 

2. V. J. Hoop's Waldhaume wnd Kulturpflanzen (Trubner, 
1905), pp. 113-114, 382-384. The question is far from being 

3. cf. G. Dottin : Les desinences verbules en r en Sanskrit en 
italique et en celtigue. He regards the passive in r in Celtic and 
Italic as an independent creation, the common element r going 
back to the period of Indo-European unity. Even the future in 
-bo he regards as a possibly analogous formation and different in 
origin and development. Compare critique in Eevue Cel'tique, 18, 
343, where M. D'Arbois de Jubainville takes exception to some 
points. Irish, contrary to the Latin, has conserved the Indo 
European perfect. Further, see G. J. Ascoli : Ossei'vazioni 
fonologiche concernenti il celtico e il neolatino in Actes du dixieme 
congres international des Orientalistes ii. eme partie, Leide Brill, 
1895; cf. Indogerm. Forschungen Anzeiger vii., i., 70. Also 
Windisch in Grwndriss der Rom. Philologie where most of the 
relative literature is summarized and discussed. The views of M. 
D'Arbois were made accessible some years ago in a paper in the 
Celtic Magazine, ed. by Dr MacBain. cf. Giles's Manual § 449. 

4. cf. Rhys's Celtae and Oalli in Proceedings of the British 
Academy. Dr MaoBain's notices of it in the Scottish Historical 
Review and in the Celtic Review are of interest, as also Sir J. 
Rhys's references in his Celtic Inscriptions of France and Italy, 
reviewed by the writer in the Scottish Historical Review, July, 

5. See Stokes on Pietish and Other Names in Bezzenherger's 
Beitrdge, Band 18. In the second edition of Skene's Highlanders 
oj Scotland, Dr MacBain clearly summarizes the whole of the 
Pietish problem. Dr Zimmer's views were made accessible in a 
paper treating of Matria/rchy Among the Picts given in the writer's 
Leabhar Nan Oleann (Edin. : N. Macleod). 

6. See Old Celtic Inscriptions by Stokes in Bezzenherger's 
Beitrdge, B. xi,. 112-141 ; Rhys's Celtic Inscriptions of France 
and Italy, and reviews by Thurneysen in Zeitschrift fur Celtiscke 


7. cf. Rhys and Jones : The Welsh People ; v. Henry's 
Lexicon Etymol., p. xxiii., where he refers to the dialects of 
Modern Breton. On the periods of Old Breton see Loth's 
Vocahulaire Vieux-Breton, Paris, 1884, ch i. 

8. The presence of z (for vowel-flanked s) can only be explained 
by assuming that the Ogmic alphabet was invented or imported 
before the regular disappearance of s between vowels — v. Bezzen- 
herger's Beifrdge, xi., 144. Mr R. A. Stewart MacAlister in his 
work on The Ogam, Inscriptions (London : D. Nutt), suggests a 
different value in the case of z ; in which case, if we have/ for z, 
we require to read v for the / of this transcription of the Ogam 

9. Add K. Meyer's Old Irish treatise on the Psalter (Oxford : 
Clarendon Press), his edition in the Revue Celtique of the Old 
Irish version of TocTimarc Emere; and Fd/ire Oengusso (2nd ed. 
by Stokes in Publications of Henry Bradshaw Society). 

10. About one half of the contents was ti:ansliterated by the 
writer in Leabhar Nan Gleann ; cf. Stern's critique in Zeitschrift 
fUr Geltische Philulogie. One of the chief poems has since been 
found in a good version in an Irish MS. from Ratisbon, of which 
lan account has been given by the writer in the forthcoming 
volume of the Transactions of the Gaelic Society of Inverness. 

11. 12, 13, 14, 15, 16, 17, 18. cf. the writer's treatment of 
The Gaelic Dialects in Zeitschrift fiir Celtische Philologie ; also 
Rev. C. Robertson on the same subject in the Celtic Review ; M. 
Maofarlane's The Phonetics of Scottish Gaelic ; and Professor 
Mackinnon on Scottish Gaelic Dialects in a paper in the Trans- 
actions of the Gaelic Society of Inverness. 

19. "Ainanlaut before a vowel seems to come from ;?. So 
apparently in Irish haue = ttciis, and Heriu cognate with TriepCa. 
This change is regular in Armenian, see Brugmann's Grundriss, 
§ 30 " — Stokes in Bezzenberger's Beitrdge, 23, 44. In last ed. of 
the FMire Stokes regards ire as the cognate of the Greek word 
cited. But this does not aiFect the cases in which an historic h 
seems to represent a vanished p ; compare the m, for n in the 
derivation of amharus ; and see Dr Pedersen's Vergleichende 
Grammatih der Keltischen Sprachen, as well as the second edition 
of Brugmann's Grundriss der Vergl. Grammatik. 

20. A great levelling, as compared with what one must infer 
from the historic development of Indo-European, has taken place 
in Gadelic. Dr MacBain's Indo-European Alphabet is therefore 
simplified in the gutturals, although perhaps it would have been 
more regular to have put in a labio-velar series 'apart. Osthoff 
recognises three k-rows, labio-velar, velar, palatal," in the mother 


speech ; v. Indogerm. Forschtmgen, 4, 264 ; Wharton's Etyma 
Latina recognise the three rows c, h, q ; cf. Zupitza's treatment 
of the gutturals. In Gadelio the velar and the palatal series have 
fallen together, but there is a distinct treatment of the labio- 

21. Contamination may have been at work here. But 
although the Cymric cognate is daigr, and Old Latin shows 
dacruma, H. German, zahar, 0. Icelandic, tdr, Germ., zahre, in 
view of the Gadelic forms, we may take the pre-historic form to. 
have been *dnkru, which developed on the Brythonic side into a 
proto-Celtic *dakru. Compare Dr Walde's Lateinisches Etymolo- 
gisches Worterbuch, p. 319, also p. 5, where L. acer is given as 
cognate with Irish Gadelic ^r, high. 

22. m^ith should be meith, as in the Dictionary, with long 
open e ; this is diphthongized in the Northern dialect as rmatfi— a 
case of diphthongization of long open e where there has been no 
compensatory lengthening. 

23. See Zeitschrift fiir Geltische Philologie, Band 3, 264, 275, 

24. See Zupitza on i, j in Celtic, in Zeitschrift fur Geltische 
Philologie, 2, 189-192. 

25. See Foy in Indogerm. Forschungen, 6, 337, on Celtic ar, 
al = Indogerm. f, I ; and Zupitza on r, I in Celtic, in Kuhn's 
Zeitschrift, 35, 253. 


Page xxxiii. — In the third line from the bottom of the page, 
for krid on, read kridion ; in the eleventh line from the bottom of 
the page, for the word in brackets, read (ballons). 

Demy 8vo, Cloth / over 400 pages. 
lOs 6d to Subscriber*; IZs 6d on Fublieation. 






This work, which has been out of print since 15 months after 
its first publication in 1896, and consequently in great demand, 
is in course of preparation for the issue of a Second Edition, in 
which Dr MacBain's corrections and additions are being embodied. 

The Dictionary deals with about 7000 of the primary words 
only ; tracing these, where possible, to their Aryan roots ; giving 
parallel words in the other Celtic Languages where such exist ; 
and also cognate forms in other European Languages and in 

No student of the Gaelic Language should be without a copy ; 
and all writers of the language would do well to possess a copy 
of the only available authority to which to refer for the correct 
forms of words. 

For reasons that can be appreciated with respect to Gaelic 
publications, the edition, now to be issued, will be strictly limited, 
and all who desire to possess copies should lose no time in placing 
their orders with the Publisher, 


Bookseller and Publisher, 
















The following Etymology of the Principal Gaelic ISTational 
Names, Personal Names, and Surnames was originally, and 
still is, part of the Gaelic EtymologicaJ Dictionary by the late 
Dr MacBain. The Disquisition on Ptolemy's Geography of 
Scotland first appeared in the Transactions of the Gaelic 
Society of Inverness, and, later, as a pamphlet. 

The Publisher feels sure that the issue of these Treatises 
in their present foim will confer a boon on those who cannot 
have access to them as originally published. They contain a 
great deal of information on subjects which have for long years 
interested Gaelic students and the Gaelic public, although they 
have not always properly understood them. Indeed, hereto- 
fore they have been much obscured by fanciful fallacies, which 
Dr MacBain's study and exposition will go a long way to 








Albion, Great Britain in the Greek writers, Gr. "AXfSiov, AX^iotv, 
Ptolemy's AXovlwv, Lat. Albion (Pliny), G. Alba, g. Albainn, 
Scotland, Ir., E. Ir. Alba, Alban, W. Alban : * Albion- (Stokes), 
" white-land " ; Lat. albus, white ; Gr. dA</)os, white leprosy, 
white (Hes.) ; 0. H. G. albiz, swan. 

Abmoric, belonging to Brittany, Lat. (Csesar) Armoricus, Aremori- 
cus (Orosius), *are-mori, " by the sea " (see air and muir in 
Diet.), M. Br. Arinory, Brittany, armor, land by the sea, Br. 
arvor, maritime. 

Britain, G. Breatann, Ir. Breatain, E. Ir. Bretan, n. pi. Bretain, 
the Britons, W. Brython, Briton, Corn. Brethon, Br. Breiz, 
Brittany, Lat. Brittania (Csesar), Brittani, Britons, B/jexTai/ot 
(Strabo). The best Gr. forms are TLpenavoi, TLpiTTaviKrj, 
W. Prydain, Britain, E. Ir. Cruithne, a Pict, 0. Ir. (Lat.) 
Gruithnii (Adamnan, Cruthini Populi) : *Qrtanid, root qrt, 
to which Stokes refers G. cruithneachd, wheat, thotigh the 
usual reference is to G. cruth, picture, form, still retaining 
the notion of " pictured " men as in the old explanations of 
Pict. Stokes, Rhys, etc., regard the Lat. Brittania as a word 
of different origin from the Gr. UpeTravia, and G. Cruithne ; 
though, as a matter of fact, the Lat. seems to have been a 
bad rendering of the Greek. The Cruithne or Picts thus 
gave their name to Britain, as being, about 300 B.C., its then 
Celtic inhabitants. 

Brittany ; the Behton language ; from Britain above. Britons 
poured into France in the fifth and sixth centuries. 

Caledonia, northern Scotland (Tacitus), Gr. KaXrjSovioi, (Ptol., 
etc.), Lat. Caledonii (Lucan, Martial, etc.), 0. G. Dun- 
Callden, Dnm-CaUen, Dwx-Keld, fort of the Caledonians, 
G. Dun-Chaillinn ; explained by Windisch as from *cald, the 
root of G. coille, the force being " wood-landers." Stokes and 
others object because of the ij (Lat. e) in KaAiyS- ; but if the 
Eng. and Gaelic modern forms are the descendants of the 
word Caledonia as locally spoken, the objection cannot hold. 

Celts, Lat. Celtx (Caesar), Gr. KeXrot, KeAraj, KeAtikos, appearing 
in the fifth and fourth cent. B.C. in Herodotus, Xenophon, 
etc. : *Kelto-s, " the lofty," root qel, raise, go, Lat. celsus. 


high, Eng tx-zel Lit. Ultas, raised. Rhys refers the name 
to the root qel, slay, Ag. S. hild, war, Norse hildr, Lat. 
percello, hit, Lit, halti, strike : the Celtse being " smiters." 

COBNWALL : Cornish, Ag. S. Gornwalas, the Walas or Welsh of 
the Corn or Horn, E. Ir. i tirib Bretann Cornn (Corm.), in 
the lands of the Britons of the Corn. For Walas see Wales. 

Cbuithnb, a Plot ; see under Britain. 

Cymry, the Welsh (pi.), Cymraeg, the Welsh name for the Welsh 
language ; the singular of Cymry is Cymro, older Gym-mro : 
*Com-mrox, pi. Com-mroges or Combroges (cf. Csesar's AUo- 
broges, "Other-landers"), country-men, "co-landers," from 
brog,' mrog of brugh in Diet., q.v. The E. Ir. Gaelic for 
Wales is found in the phrase isinchomreic = im Kymrischen 
(Zim. Zeit. ^2 162). 

Erin ; see Ireland. 

Gaelic, Gael, the name of the language and people of the Scottish 
Highlands, G. Gaidhlig, G^idheal, Ir. Gaoidhilig, Gaedhilig, 
the Irish language, Gaoidheal, Irishman, E. Ir. Gdedel (1100 
A.D.), Gaideli (Giraldus), W. Gwyddel, Irishman : *Gddelo-s 
(for So. Gaelic) or *Gdidelo-s (for Irish), root ghddh, Eng. good, 
Gar. gut, etc. 1 The Scotch form seems the best, as its use 
has been continuous, the race being only a fourth item in 
Scotland. Stokes gives a pro to-Gaelic *Goidelos or *Geidelos, 
which Bez. compares to the Gaul. Geidumni, and which 
Stokes compares with Lat. hoedus, goat (" Goat-men," cf. 
Oscan Hirpini) or Lit. gaidys, cock. 

Galli, Gaul, now France, Lat. Gallus, Galli (fourth to first cent. 
B.C.), Gr. raXctTTjs, VaXdrai (third and second cent. B.C.) ; 
from the root gal, bravery, which see in Diet., with discussion 
of Galli and G. Gall, Lowlander, stranger. 

Ikeland, Irish ; G. 'Eireann, Ir. 'Eire, g. 'Eireann, E. Ir. 'Eriu, 
'Erenn, W. Twerddon, Iwerddon, M. W. Ewyrdonic, Irish, 
Ptol. 'lovfpvla 'lepvq (Strabo), Lat. Hibernia, Iverna (Mela), 
leme (Claudian, fourth cent. A.D.), Evernili, Irish (Adamnan) : 
*Iverj6n-, *Everj6n-, usually referred to Piverjo-, Skr. pivari, 
fat, Gr. TLiepia, the Grecian seat of the Muses, irimv, fat 
(Windisch, Stokes): "rich-soiled, swelling." Others refer it 
to G. iar, west, or Skr. dvara (froni ava, G. bho), western, 
lower. No derivation can be satisfactory which does not at 
the same time account for the similarly named Highland 
rivers called 'Eire, 'Eireann, Eng. Earn, Findhorn. 

Man, Manx; Manx Manninagh, Manx (adj.), Gailch, Gaelk, the 
Manx Gaelic, E. Ir. inis Manann, Isle of Man, a genitive from 
*Mana ( = Lat. Mono), early W. Manau, Lat. Mona (Csesar), 


Ptol. MovdoiSa, Monapia (or Afona ?) The E. Ir. god-name 
Mananndn Mac Lir (son of the Sea) is connected with 
the Island ; Skr. Manu, the Law-giver ; Teutonic Mannus 
(Tacitus), Eng. man. 

PiCTS ; G. Cruithnich, for which name see under Britain. The 
name Picti can scarcely be separated from the Gaul. Pictavi, 
now Poitiers ; and, if this be the case, the usual derivation 
from Lat. pictus, painted, must be abandoned. Windisch 
adduces E. Ir. cicht, engraver, carver, for which a Brittonio 
piht, pict may be claimed as a parallel (*qict) ; this again 
leaves the idea of tattooing intact, and so agrees with the 
historical facts. 

Scotland, Scots ; E. Ir. Scott, pi. n. Scuit, d. Scottaib, Irishmen ; 
Adamnan — Scotia, Ireland, Scoti, the Irish, Scoti Britanniae, 
Scots of Dalriada, etc., Scoticus, Irish, Scotice, in the Gaelic 
language, Lat. (fourth cent.) Scotti, Scdti, *Skotto-s. Stokes 
translates the name as "masters, owners," allied to Got. 
skatts, money, Ger. schatz, treasure, stock, Oh. SI. skotii, pro- 
perty, cattle. The root skat, hurt, scathe, cut, of Eng. 
scathe, has been suggested, either as " cutters" or " tattooed 
ones" (so Isidore of Saville). Rhys has suggested connection 
with W. ysgwthr, a cutting, carving — " tattooed or painted 

Wales, Welsh ; Ag. S. Wealas, Walas, the Welsh — the name of 
the people in pi. being used for the country, Wylisc, Welsh, 
Wylisce men, the Welsh ; sing, of Wealas is Wealh, a foreigner, 
Welshman, 0. H. G. walk, foreigner, Gelt, Ger. wal- in wal- 
nuss, Eng. wal-nut : from the Gaul, nation of the Volcae, 
bordering on the Germans, *Volko-s, *VolM, "the bathers," 
from vole, bathe (see failc in Diet.). Stokes connects the 
name with Lit. willeti, pull, referring to the restless wander- 
ings of the Gauls. 


Adam, G. Adhamh, Ahu (Fer. MS.), Awzoe (D. of L.), E. Ir. Adam, 
0. Ir. Adim (g) ; from Hebrew Adam, red. Hence Mac- 
adam, M'Caw, and from Dial. G. 'Adaidh (a diminutive from 
So.) M'Cadie, M'' Adaidh. 

Adamnan, G. Adhmbnan (pronounced Yownan or Yonan), earlier 
Adhamhnan {Ogharnhnan, M'V.), E. Ir. Adamndn, Lat. 
Adamnanus (seventh cent.), St Adamnan (died 704 a.d.), 
" little Adam," a Gaelic diminutive from .4 rform. Hence the 
personal name Gilleownan (1495), Giolla-Adkamhndin, father 
of Somerled (twelfth cent.), Qilla-agamnan (1467 MS.), 
whence Skene deduces the Mac-lennans, q.v. 

Albxandek, G. Alasdair, Allex'' (D. of L.), Alaxandair, (1467 
MS.), M. Ir. Alaxandair; from Lat. Alexander, from Gr. 
AXe^avSpos, " defending men." Hence G. M'Alasdair, Mac- 
alister ; further Mac-andie (from Sandy). 

Allan, G, Ailean, E. Ir. AiUne, Adamnan's Ailenus, from al, 
rock '? The Norman Alan, whence Scotch Allan mostly, is 
0. Br. Alan, Alamnus, Nennius Alanus, from Alemannus, the 
German tribe name — "All Men." Of. Norman, Frank, 
Dugall, Fingall. Hence Mac-allau. 

Alpin, G. Ailpein, E. Ir. Alpin (Dalriadic king 693) ; from Pictish 
or Welsh sources — M. W. Elphin, Elfin, which Stokes sug- 
gests to be from Lat. AlMnus, from albus, white (or allied 
rather 1). Hence G. M' Ailpein, Mac-alpine. 

Andrew, G. Aindrea (Anndra, Dial.), G-illeanndrais, Eng. Oil- 
landers, St. Andrew's gille, M. G. Andro (D. of L.), Ainnrias, 
Oille-ainnrias (1467 MS.), E. Ir. Andrias ; from Lat. Andreas, 
g. Andrew, from Gr. ,'AvS/3eas, a reduced double-stemmed 
name now showing only dvSp-, man (see neart). Hence 
Mac-andrew, Gillanders, Anderson. 

Angtts, G. Aonghas, Ir. Aonghus, g. Aonghusa, E. Ir. 'Oengus, 
0. Ir. 'Oingus, W., Cor. tfngust : Oino-gustu-s, " unique 
choice," from aon and gus, choice (Eng. choose, Lat. gustus, 
taste, as in G. tagh). Hence M'Aonghuis, Mac-innes ; further 

Archibald, G. Gilleasbuig, Bishop's gille (see eashuig in Diet.), 
M. G. Oillespik (D. of L.), Gilla-espic (1467 MS.). Hence 
Gillespie. The name Archibald, Ag. S. Arcebald, Arcenbald 


•or Ercenhald, which vaguely means " right-bold" (0. H. G. 
erchen, right, real), has no apparent connection with Gillespie 
in meaning or origin (cf. similarly Ludovic and Maol- 

Arthur, G. Artair, M. G. Artuir, E. Jr. Artuir, Artur, Jr. Lat. 
Arturius, son of ^Edan (Adamnan), W. Arthur, to which the 
Lat. Artorius (Juvenal) has been compared and suggested as 
its source (it being maintained that the Gens Artoria of 
Yorkshire lasted from Roman to Domesday-BooK times, where 
Artor appears in the days of Edward the Confessor). If 
native to Brittonic (which is probable), it is from *arto-s, 
a bear, W. arth, 0. Ir. art, whence the names Art, Artgal, 
Artbran. Rhys prefers to render the *arto- as "oultor," 
from ar, plough {Arth. Lpij., 40-48), allying Arthur to the 
idea of a " Culture God." Hence 6. M' Artair, Mac-arthur. 

Bain, from G. ban, white. The Bains of Tulloch appear in the 
sixteenth century variously as Bayne or Bane, with a con- 
temporary near them called John Makferquhair S'PGillehane 
(1555). This last name is now M'Ille-bha,in, "¥si\r-gille," 
rendered into Eng. by Whyte ; whence also AJ'Gilvane. 

Eartholomew, G. Parian, Ir. Parthalon, E. Ir. Partholdn, Lat. 
Partholomceus or Bartholomceus (Nennius, ninth cent.), the 
name of a personage who is represented as the first invader 
of Ireland after the Flood (278 years after ]). The p proves 
the name to be non-Gadelic ; and as the historians take 
Partholon from Spain, the Spanish Bar Tolemon of legend 
has been suggested as the original. Prof. Rhys thought it 
came from the Ivernians or Pre-Celtic race in Ireland. Hence 
tlie Clan Mac-farlane, G. M'Pharlain. 

Brown, G. M'A'-Bhriuthainn, M. G. M'abfiriuin (1408 Gaelic 
Charter), from britheamhain, the former (Sc. Gaelic) genitive 
of britheamh, judge, q.v. Hence Mac-brayne. 

€ambeon, G. Camshron, Camaran, M. G. Cdmsroin, g. (M'V.), 
Camronaich (D. of L.), Gillacamsroin (1467 MS.), Charter 
Eng. Camroun (1472); explained as from cdm-srhn, "wij- 
nose,'' which is the most probable explanation (cf. cainibeul, 
E. Ir. eerrbdl, wry mouth). Connection with carnerarius or 
chamberlain (of Scotland) unlikely, or with the fourteenth 
century Be Cambruns or Cameron parish in Fife. 

Campbell, G. Caimbeul, M. G. Cambel (1467 MS.), Gambell 
(1266, etc.), from cambel, wry-mouthed {cam and beul ; see 
Cameron). There is no De Cambel in the mmierous early 
references, but De Campo-bello appears in 1320 as a Latin 


form and an etymology ; this, however, should naturally be 
Be Bello-eampo as Norman-French idiom and Latin demand — 
a form wo have in Beau-champ and Beecham. De Campello 
or De Campellis (little plain) has been suggested ; but 
unfortunately for these derivations the earliest forms show 
no de : Gamhell was an epithet, not a place-name. 

Caemichael, G. M'Grillemhicheil, Son of the gille of St Michael, 
M. G. Gillamichol (1467 MS.), O.G. Giliemicel (B. of Deer). 
The name Carmichael is really Lowland — from the Parish 
name of Carmichael in Lanark (Michael's caer or cathair, q.v.). 

Cattanach, Chattan, G. Catanach, M. G. plural Cattanich (D. of 
L.), "belonging to Clan Chattan," Clann Gillacatan (1467), 
which claims descent from Gillacatain (1467 MS.), servant 
of St Catan, whose name denotes "little cat" (see cat). 

Chaeles, G. Tearlach, M. Jr. Toirrdhealhhach (Maclean Gene- 
alogy), Englished as Tirlagh and Turlowgh, E. Ir. Toirdelhach, 
Latinised and explained as Turri-formis, " Tower-shaped," 
but the toir in Gaelic took the phonetics of the prefix tair, 
super, and hence the modern G. form. Hence M'Kerlie. 

Chisholm, G. Siosal, Siosalach, De Cheiholme (thirteenth century 
documents), De Cheseholme (1254), a Border name, the place- 
name Chisholm being in Roxburgh : Ghes-holm (a holm, but 
Ghes 1). 

Claek, G. Cl^ireach ; see cUireach in Diet. Also M' A'-Chl^irieh, 
whence Galwegian M'Chlery. 

Coll, G. Colla, M. G. Golla (M'V., 1467 MS.), E. Ir. Golla : *Gol- 
navo-s, from col, eel, high, as in Geltce (see above). 

Colin, G. Cailean, M. G. Gallane (D. of L.), Cailin (1467 MS.), 
Colinus (Lat. of 1292). This is a personal name, once more 
or less peculiar to the Campbells, the Chief being always in 
Gaelic M'Cailein. Its relation to Eng. and Continental Golin 
is doubtful. Cf. Goiledn, " whelp," and personal name ; the 
G. is a dialectic form of old coiledn (see Fol.), cuUean, 

Ckeeae, G. Criathrar, the name of a Lochtay-side clan who- 
regard themselves as Mackintoshes, explaining the name as 
"riddler," from criathar (which see in Diet.) : the derivation 
is right, but for the meaning compare the Eng. noun and 
name Sieve{w)right. See Gelt. Mag. ^, 38. 

Gumming, G. Cuimein, Cuimeanach, earliest Eng. form Gomyn, a 
Norman family dating from the Conquest, belonging to the 
Norman house of De Comines, a territorial designation. 


David, G. Daibhidh (Classical), D^idh (C.S.) ; hence Clann Ditidh 
or the Davidsons, a branch of the Clan Chattan. [n C.S., 
Davidson appears as D^ibhiosdan. 

Dermid, G. Diarmad, M. G. Dermit (D. of L.), Diarmada, gen. 
(1467 MS.), E. Ir. Diarmait, 0. Ir. Diarmuit, Diarmit, Ir. 
Lat. Diormitius (Adamnan). Zimmer explains the name as 
Dia-ermit, " God-reverencing," from dia and ermit : *are- 
ment-, "on-minding," root ment, as in dearmad, q.v. 

Dewar, G. Deoir, Deoireach, documents Ddire (1487), Jore 
(1428); from debradh, a pilgrim, q.v. Hence Macindeor. 

Donald, G. Domhnall, M. G. Domnall (1467 MS.), gen. Donil 
(D. of L.), 0. G. Domnall (B. of Deer), E. Ir. Domnall, Ir. 
Lat. Domnallus (Adamnan), Domnail (do., ablative), Early W. 
Dumngual, later Dyfnwal : * Dumnovalo-s, from dubno- of 
domhan, and valo- (seeflath), meaning " world-wielder, world- 
ruler," much the same in meaning as Dumnorix, world-king, 
Caesar's opponent among the Aedui. See domhan, flaih. 
Hence M'Dhdmhnuill, Mac-donald. 

Duff, M. Ir. Duhh {Clann Duhh, Clan DufF, of which was Mac- 
beth, etc.), earlier Dub, King DuiF in tenth century ; from 
Gadelic dub, now duhh, black, q.v. As a personal name, it is 
a curtailment of some longer or double-stemmed name (cf. 
Fionn, Flann, red). Hence Macduff (Clen m'^ Duffe, 1384). 
The family name D^iff is merely the adjective duhh used 

Duffy, Ir. Dubhthaigh ; see Mac-phee. 

DuGALD, G. Diighall, M. G. Dowgall, g. Dowle (D. of L.), Dubgaill, 
gen. (1467 MS.), thirteenth century documents give Dugald 
(1289), Dufgal (12&1), M. Ir. Dubgall (first recorded Dubgall 
is at 912 A.D.), from Early Ir. Dubgall, a Dane, "Black 
stranger," as opposed to Finngall, a Norwegian, " Fair 
foreigner." See, for derivation, jionn and Gall. Hence 
M'Dhughaill, Mac-dougall, Mac-dowel, etc 

Duncan, G. Donnchadh (Dial. Donnach), M. G. Duncha (D. of L.), 
Donnchaid, gen. (1467 MS.), 0. G. Donchad (B. of Deer), 
E. Ir. Donnchad: * Donno-catu-s, * Dunno-catu-s, "Brown 
warrior," from donn and cath, q.v. The Gaulish Donna- of 
personal names has been referred by De Jubainville to the 
same meaning and origin as M. Ir. donn, king, judge, noble — 
a word occurring in O'Davoren's glossary. 

Edward, G. 'Eideard ('Eudard, Dial), Imhear, lomhar; the first 
is the Eng. Edward borrowed, the second is the Norse Ivarr 
borrowed (see Mac-iver). Hence M"Eideard, M'Fdward. 


EwEN, G. Eoghann (Dial. Edghalnn), M. G. Eogan, Eoghan, E. Ir., 
0. Ir. Eogan : *Avi-gono-s {*Avigenos, Stokes), '•' well born, 
good," from *avi, friendly, good, Skr. dvi (do,). Got. avi-liud, 
thanks, Lat. aveo, desire, possibly Gr. ei'-, good (cf. here 
Ei'yei'ijs, Eugenius), W. has Eu-Ugirn, Etirtut, 0. Br. Eu-cant, 
Eu-hocar, Gaul. Avi-cantus. Rhys {Hib. Led. 63) refers Ir. 
Eoghan and W. Owen to *Esu-gen-, Gaul. Esugenus, sprung 
from the god Esus. Zimmer regards Omen as borrowed from 
Lat. Eugenius. Cf,, however, the evo- of Ogmic Eva-cattos, 
now Eochaidh. Hence Mac-ewen. 

Farquhar, G. Fearchar, M. G. Fear char, Fear chair, Ir. Fearchair 
(F. M., year 848 a.d.) : *Ver-caro-s, "super-dear one"; for 
fear, see Fergus, and for car see Diet, above. Hence 
MThearchair, Mac-erchar, Farquharson, M'Farquhar. 

Fergus, G. Fearghas, M. G. Fearghus, Fergus, E. Ir., 0. Ir. 
Fergus, g. Fergusso, W. Gurgust, 0. Br. Uuorgost, Uurgost : 
*Ver-gustu-s, "super-choice"; for ver- ox fear-, see in Diet. /ar, 
air (allied to Lat, super), and for gustus, see under Aonghus 
above. Some regard Fer here as G. fear, man, *viro- or *vir. 

FiNGAL, G. Fionn, Macpherson's Gaelic Fionnghal, which really 
should mean " Norseman," or Fair-foreigner, M. G. Fionn- 
ghall, a Norseman (M'V.), ri Fionn-gal, king of Man and the 
Isles (M'V.), Fingal {Manx Chron.\ king of Man and the 
Isles from 1070 to 1077 : from jionn and Gall, q.v. Fingal 
as the name of the Gaelic mythic hero is an invention of 
Macpherson's, as likewise is his Gaelic Fionng^hal. As a 
matter of fact the name is a Gaelic form of the female name 
Flora ! See Fionnaghal in the addendum to this list. 

FiNLAY, G. Fionnla, Fionnlagh (misspelt Fionnladh), M. G. 
Finlay (D. of L.), Finlaeic, gen. (1467 MS.), Fionnlaoich, 
gen. {Duan Albanach), E. Ir. Findlcech (Lib. Leinster), Finn- 
loech and Finlaeg, gen. (Marianus Scotus). Those early 
forms and the Norse Finnleihr prove that the name means 
"Fair hero" (fionn and laoeh). It is a popular (10th and 
11th century) rendering of Finnlug, "Fair attractive one," 
the older name. It has been explained as " Fair calf," which 
would suit the phonetics also. Hence Finlayson, Machinlay 

Forbes, G. Poirbeis, Foirbeiseach, early document form De Forbes 
(thirteenth cent.), so named from the place-name Forbes in 

Eraser, G. Friseal, Frisealaoh, circ. 1298 the patriot's name is 
variously Simon Eraser, Frasel, Fresel, Frisel, in Domesday 
B. Fresle, Battle Abbey Eolls (?) Frisell or Fresell ; usually 


referred to 0. Fr. freze, a strawberry, *frezele, from Lat. 
fragula, fragum, Fr. fraisier, strawberry plant. For sense, 
of. the name Plantagenet (broom). Strawberry leaves form 
part of the Fraser armorial bearings. The word may also 
mean "curled" (^^g. frizzle, frieze). 

Galbraith, G. M' A'-Bhreatnaich, son of the Briton (of Strath- 
clyde). The name appears in the thirteenth century in 
Lennox, etc., as Galbrait (from Gall and Breat- of Breatann 

George, G. Se6ras, Sedrsa, Deorsa, ultimately from Gr. yeu/iyds, 
a farmer, "worker of the earth" (y-ij, earth, dpyos, Eng. 
work). Hence the Border M^George. 

Gilbert, G. Gilleabart, Gillebride. Gilbert is from Ag. S. Gisle- 
hert, " Bright hostage " (see giall in Diet.); Gillebride is St 
Bridget's slave, an exceedingly common name once, but now 
little used. 

Gilchrist, G. Gillecriosd, M. G. Gillacrist, Ir. Gillacrist (several 
in eleventh century): "servant of Christ." Hence M' Gil- 
christ. It translates also Christopher. 

Gillespie, G. Gilleasbuig ; see Archibald. 

Gillies, G. Gilliosa : " servant of Jesus." From M'A-Lios 
comes the " English " form Lees, M'Leish. 

Glass, G. Glas, an epithet, being glas, grey. See M^Glashan. 

Godfrey, G. Goraidh, M. G. Gofraig (1467 MS.), Godfrey (do.), 
Ir. Gofraidh (F.M.), M. Ir. Gothfrith, Gofraig, also Gofraig 
(Tigernach, 989), E. Ir. Gothfraid (Lib. Lein.), E. W. Gothrit 
{Ann. Camb.). The Norse name, for it is Norse-men that are 
referred to, is Go&roSr or Gudrod (also G6r6&r), but the 
earlier Gaelic shows rather- a name allied to the Ag. S. 
Godefrid, Ger. Gottfried, " God's peace." Modern Gaelic is 
more like the Norse. The Dictionaries give G. Guaidhre as 
the equivalent of Godfrey; for which, however, see M'Quarrie. 

Gordon, G. G6rdan, Gordon, Gordonach ; from the parish name 
of Gordon in Berwickshire. The De Gordons are well in 
evidence in the thirteenth century. Chalmers explains the 
place-name as Gor-dyn, "super-dunum" (see far and diin). 

Gow, G. Gobha, a smith, now usually gobhainn, q.v. Hence 
Mac-cowan, Mac-gowan, Cowan. 

Grant, G. Grannd, Grant (1258), an English family which settled 
about Inverness in the thirteenth century, Eng. Grant, 
Grand, from Fr., Eng. grand. 

Gregoe, G. Griogair, Griogarach, M. G. M'Gregar (D. of L.), 
M. Ir. Grigoir, E. Ir. (Lat.) Grigorius (Gregory the Great, 
died 604), from Lat. Gregorius, Gr. Fpi^ydpios, a favourite 


ecclesiastical name from the third century onward (of. Gr. 
yp-qyopiia, be watchful, Eng. care). Hence M'Griogair, Mac- 
gregor, Gregory. 

GuNN, G. G-uinne, G-unnach, early documents Gun (1601), Glan- 
gwn (1525), in Kildonan of Sutherland, originally from 
Caithness ; from the Norse Gunni (twelfth century), the 
name then of a son of Olaf, a Caithness chief {Ork. Saga). 
This Gunni is a short or "pet" form of some longer name of 
two stems, with gunn-r, war, as the first and chief one (cf. 
Gann-arr, which is an old Orkney name, Gunn-hjorn, Gunn- 
laugr, Gunn-dl/r, war-wolf, Gunn-stein, Gunn-valdr). 

Harold, G. Harailt, M. Ir. Aralt, from Norse Haraldr (same in 
roots and origin as Eng. herald). Hence Mac-raild. 

Hector, G. Eachunn (Dial. Eachainn), M. G. Eachuinn, g. (M'V.), 
Eachdhuin, g. (M'V.), Eachdhqnn, g. Eachduinn (1467 MS.), 
Ir. Eachdonn (year 1042): *Eqo-donno-s, "horse lord," like 
Each-ihighearna of Mac-echern. Of course " Brown-horse " is 
possible ; cf. Gr. Sav6t7r?ros. The phonetics are against 
*Each-duine, "horse-man," as an explanation. 

Henry, G. Eanruig'; from 0. Eng. Henric, now Henry, from 
Germanic lieim-rik, " home-ruler " (Eng. home and ric in 
bishop-ric, rich). Hence Maclcendrich, Henderson. 

Hugh, G. 'Uisdean (Huisdean), in Argyle Eoghan, M. G. 
Huisdumn, which comes from Norse Eysteinn, "jS^y(?)-stone." 
The Dictionaries also give the G. Aodh (see Mackay) as 
equivalent to Hugh, which is itself from Germanic sources, 
Teutonic root hug, thought. ■ 

Jambs, G. Seumas, M. G. Semus (M'V.) ; from the Eng. James, a 
modification of Hebrew Jacob. 

John, G. Iain, older E6in, in compounds Seathain, as Mac-Gille- 
Sheathainn, now M'llleathainn. 

Kathel, G. Cathal, M. G. Cathal (M'V.), Ir. Gathal (common 
from seventh century onwards), 0. W. Catgual : *Katu-valo-s ; 
see cath, war, and val under Donald. Hence M^All, MacJcail. 

Kennedy, G. Ceanaideach, Ceanadaidh, Kennedy {Kenedy, John 
M'Kennedy, fourteenth century) is the family name of the 
old Earls of Carrick, now represented by the Marquis of 
Ailsa ; it is a famous Irish name borne by the father of Brian 
Boru in the tenth century — Ir. Ceinneidigh, E. Ir. Cennetich, 
gen. ; from ceann, head, and eitigh, ugly : "ugly head." Called 
also M'Ualraig from Walrick Kennedj^ (sixteenth century), 
who first settled in Lochaber : Walrick may be G. Ualgharg 
confused with Teutonic Vlrick, older Uodalrich, " rich 


Kenneth, G. Coinneach, M. G. Coinndech, Goinnidh, g. Coinndigh, 
g. (M'V.), 0. G. Gainnech, g. Gaennig (B. of Deer), E. Ir. 
Gainnig, gen., Ir. Lat. Gainnechus (Adamnan) : *Gannico-s, 
" fair one," from the same stem as cannach (root qas), q.v. 
The Eng. Kenneth is a different word : it is the old Scotch 
king name Cinced (E. Ir. form), 0. 6. Ginathd (B. of Deer), 
Ir. Ginaedh, " fire-sprung," from cin of cinn and aed of 

Lachlan, G. Lachlann (Dial. Lachlainn), Lachunn, M. G. Loch- 
linn, g. (M'V.), Lochloinn, n. and g., Lachlan, g. (1467 MS.), 
Ir. Lochlainn Mac Lochlainn (F.M., year 1060) ; probably 
from Lochlann, Scandinavia, possibly commencing as Mac- 
Lochlainne, a Scandinavian (" son of L."). Lochlann 
evidently means " Fjord-land." 

Lamosd, G. M'Laomuinn, Laman, M. G. Ladmann, early docu- 
ments Lawemiindus (Lat. of 1292), Laumun (circ. 1230), 
M. Ir. Laghmand, Lagmand ; from Norse lagaina&r, logmaS'r, 
lawman, pi. I'dgmenn, "law-men," by meaning and derivation. 
Hence M'Glymont (D. of L.), VGlymont, Clyne lymyn. 

Laurence, G. Labhruinn, M. G. Labhran (1467), Ir. Laurint 
(Saint), from Lat. Laurentius, St Laurence, the ultimate stem 
being that of Lat. laurus, a laurel. Hence M'Labhruinn, or 

Lewis, G. Luthais ; from Fr. Louis, from Ghlovis, the Frankish 
king (fifth century), degraded from old German Ghlodwig, 
now Ludioig {* Kluto-vigo-s, famed warrior, roots in cliU and 
Eng. victory). Hence Eng. Ludovic, which is rendered in 
G. by Maold6nuich, shaveling of the Church. 

Livingstone, G. M'An-l^igh ; see Mac-leay. 

Luke, G. Lucais. Hence Mac-lucas. 

Magnus, G. Manus, MAnus, M. G. Magnus, Manuis, g. (1467 
MS.), Ir. Maghnus, Norse Magnilss, from Lat. magnus, in the 
name of Gh&xlemagne — Carolus Magnus. 

Malcolm, G. Calum, earlier G-illecalum, M. G. Mylcollum (D. of 
L.), Maelcolaim, 0. G. Malcoloum, Malcolum, Gilliecolaim, 
(r. Maelcoluim : from maol, bald, and calum, a dove (Lat. 
columba), the particular Galum meant here being St Oolumba. 
Hence Maccallum. 

Malisb, G. Maoliosa, E. Ir. Maelisu, servant of Jesus. Hence 
also Mellis. 

Mathbson, G. M'Mhathan, Mathanach, M. G. Mac-Matgamna 
(1467 MS.), Macmaghan {Exchequer Rolls for 1264), the Ir. 
Mac-maJion, "son of the bear," for which see mathghamhuin. 
Matheson in Perthshire and Kintyre is, as elsewhere outside 
the Highlands, for Mathew-son, G. M'Mhatha. 


Mknzies, G. M6innear, M6inn and Mfeinnearach locally, early 
documents de Mengues (1487), de Meyners (1249); De 
Meyneria would mean much the same as Be Camera, that is, 
" of the household," from me&n-, masn-, giving Fr. m^n- (our 
manage, menagerie, menial), from Lat. mans- (our mansion), 
from maneo, remain. The root anyway is man of mansion 
and manor, and the name is allied to Manners and Main- 

MoEGAN, M. G. Clann Mhorguinn (M'V.), 0. G. Morgunn, g. 
Morcunt, W. Morgan, Cor. and 0. Br. Morcant : Mori-canto-s, 
"sea-white," from the stem of muir and root knd, burn, as in 
connadh (Lat. candeo, shine, Eng. candle). See Mackay. 

Morrison, G. Moireasdan, earlier M'Grille-mhoire, Mary's servant, 
M. G. Oillamure, whence Gilmour. The name Morris is for 
Maurice, from the Latin saint's name Mauricitis, "Moorish." 

MuNRO, G. Rothach, Mac-an-Rothaich (Dial. Munro). In the 
fourteenth century the name is " of Monro," which shows it 
is a territorial name, explained as Bun-roe, the mouth of the 
Eoe, a river in County Derry, Ireland, whence the family are 
represented as having come in the eleventh century. 

Murdoch, G. Mulreach, Murchadh ; the first is M. G. Muiredh- 
aigh, gen. (M'V.), Murreich (D. of L.), Muireadhaigh, g. 
(1467 MS.), Ir. Muireadhach, E. Ir. Muiredach, 0. Ir. (Lat.) 
Muirethachus, Adamnan's Muiredachus, " lord," allied to 
muirenn and muriucdn ; Ag. S. masre, clarus ; Br. cono- 
morios C?) (Stokes R. C. 1876.) The form Murchadh is in Ir. 
the same, E. Ir. Murchad : * Mori-catu-s, sea warrior. Hence 
(from the first) M'Mhuirich (in Arran, etc., becoming Currie}, 
and from the second, Murchison, Murchie, and Ir. Murphy. 
See murrach above. 

Murray, G. Moirreach ; from the county name Moray or Murray, 
early Gadelic forms being Moreb, Muref, and Norse Morhaefi 
(influenced by Norse haf, sea) : *Mor-apia, from mor of muir, 
sea, and *'apia, the termination pf several Celtic place-names- 
Andrew Morrich, Kiltearn, 1672. 

Myles, G. Maolmoire, servant of Mary, an old and common name. 
Myles is from the Med. Lat. Milo, with a leaning on miles^ 
soldier — a common name in the JMiddle Ages. 

Mac-alister ; see Alexander. 

Mac-andbbw ; see Andrew. 

Mac-arthur ; see Arthur. 

Mao-askill, G. M'Asgaill ; from Norse 'Askell, for *'As-ketill, the 
kettle (sacrificial vessel) of the Anses or gods : " a vessel of 


Mac-aulay, G. M'Amhlaidh, Ir. Mac Amhlaoibh, M. Ir. Amlaibh, 
E. Ir. Amldih, 'Alaib ; from Norse 'Oldfr, Anlaf (on coins), 
"the Anses' relic" (Eng. Ze/i). 

Mac-bean, G. M'Bheathain, from Beathan, Englished as Bean 
(1490, Beane, 1481) or Benjamin : *Bitdtagno-s, life's son, 
from beatha, life, with the termination -agno-s, meaning 
" descendant of," Eng. -ing, now used like the Eng. to form 
diminutives. Also Mac-bain, Mac-vean. 

Mac-bbth, G. M'Bheatha (Dial. M'Bheathain and M'Bheathaig), 
M. G. Macbethad, 0. G. Mac-bead (B. of Deer), M. Ir. Mac- 
bethad, Macbeth 1058, 1041 ad.) : "son of life," from beatha, 
life. It is a personal name originally, not patronymic. 
From Macbeth come M'Bey, M'Vey, M'Veagh. 

Mac-caig, G. M'Caog, Ir. Mac Taidhg, son of Teague, E. Ir. Tadg, 
possibly allied to Gaul. Tasgius, etc. Tadg explained by 0. 
CI. and Dav. as "poet." 

Mac-callum, G. M'Caluim ; see imder Malcolm. 

Mac-codeum, G. M'Codrum ; from Norse Guttormr, GoSormr, 
Ag. S. Guthrum : "good or god serpent" (orm). 

Mac-coll, G. M'CoUa ; see Coll. 

Mac-combie, G. M'Comaidh, M. G. M'Comie (D. of L.) : "son of 
Tommie," or Thomas. 

Mac-cx)nachie, G. M'Dhonnchaidh, son of Duncan, which see. 
The Clan Donnachie are the Robertsons of Athole, so-named 
from Duncan de Atholia in Bruce's time : the English form 
of the name is from Eobert, Duncan's great-grandson, who 
helped in bringing the murderers of James I. to execution. 

Mac-cobmic, G. M'Cormaig, from Cormac (Cormag), E. Ir. 
C'ormac, Adamnan's Cormacus : *Corb-mac, charioteer, from 
corb, chariot, Lat. corbis, basket. See carbad. From corb also 
comes Cairbre, 0. Ir. Coirbre. 

Mac-coequodalb, M'Corcadail, M. G. Gorgitill, g. (D. of L.), early 
documents Mahcorquydill (1434); from Norse Thorketill, 
Thor's kettle or holy vessel (see Mac-ashill). 

Mac-ceimmon, G. M'Cruimein ; from Rumun (on a Manx Rune 
inscription), from Norse Hromundr (for Hr6&-mundr, famed 
protector) ? Ceannfaelad Mac JRumain, Bishop, d. 820 ; 
Euman, the poet, d. 742; Buman, the bishop, d. 919. Erig 
a n-agaid Rumuind, MS. Bodl. Lib. Laud. 610, fol. 10, a, a 
(0. Don's Gram.). 

Mac-culloch, G. M'Cullach, early documents M'Culloch (1458), 
M'Cullo, M'Cullach (1431)— in Easter Ross: "son of the 
Boar " (cullach) 1 M'Lulach, son of Lulach (little calf f), has 
been suggested, and this appears as M'Lulich. 


Mac-dermid ; see Derinid. 

Mac-don ALD ; see Donald. 

Macduff ; see Duff. 

Mac-bchern, G. M'Eachairn, M. G. M'Caychim (D. of L.), early 
documents Mackauchern (1499), Ir. Echthighern (Annals 
846 A.D.) : " Horse-lord," from each and tighearna. Also 
Englished as M'Kechnie (^MacEchthigerna). 

Mac-fadyen, G. M'Phaidein, early documents M'Fadzeane (1540); 
from Paidean, Pat, a pet form of Patrick. 

Mac-farlanb ; see Bartholomew. 

Mac-gill ; from a G. M'Grille, used as a curtailment, especially of 
Mac-millan or M'Gille-mhaoil. 

Mac-gillivray, G. M'Gillebhr^th, son of the Servant of Judg- 
ment, from brath, judgment, q.v. 

Mac-glasiian, G. M'Glaisein, a side-form of M'G-hilleghlais, the 
Grey lad, M. G. M'lllezlass (D. of L.), documents M'Gille- 
glasch (1508). For the formation of this name, cf. G-ille- 
naomh (Mac-niven), Gille-maol (Mac-millan), M^Gillebane 
(1555), WGille-uidhir (M'Clure, dun lad), Gilroy, red lad. 

Mac-gowan ; see under Gow. 

Mac-gregor ; see Gregor. 

Mac-hardy, G. M'Cardaidh : 

Mac-indboe ; see Dewar. 

Mac-innbs ; see Angus. 

Mac-intyee, G. Mac-an-t-saoir, son of the carpenter ; see saor. 

Mac-iver, G. M'lamhair, M. G. M'Imhair (1467 MS.), Ir. Imhar, 
E. Ir. Imair, g. ; from Norse 'Ivarr. 

Mackay, G. M'Aoidh, from Aoldh, 0. G. Aed, 0. Ir. Aed, Adam- 
nan's Aldus, g. Aido : *AidvrS, fire, E. Ir. aed, fire, Gr. aidos, 
fire, brand, Lat. aedes, house ( = hearth), aestus, heat, 0. H. G. 
eit, fire, pyre. Hence the Gaul. Aedui. 

Mac-kbllar, G. M'Ealair, M'Eallair, old documents Makkellar 
(1518), Makalere (1476), M'Callar (1470), all "of Ardare" 
in Glassary, Argyle. Ellar M'Kellar (1595), proves the 
name to be Ealair. M. Ir. Elair, the Gaelic form of 
Lat. Hilarius borrowed. 

Mac-kenzie, G. M'Coinnich ; from Goinneach, which see under 

Mackbrchae, G. M'Phearchair ; see Farquhar. 

Mackessack, for G. M'Isaac, son of Isaac. Also Mackieson, 
M'Keseh, 1475 ; Kessokissone, Kessoksone, 1488 ; Makesone, 
1507 ; Makysonn, 1400 (mostly in Menteith and S. Perth), 
from Kessoc, Kessan, personal names circ. 1500, also St. 
Kessog or Kessock. 


MickiLLOP, G. MThilib, for Philip { = Filip), where / {=ph) is 
aspirated and disappears ; from Lat. Phillipus, from Gr. 
^'tAiTTTTos, lover of horses (see gaol and each). 

Mackinlay, G. M'Fhionnla(idh) ; "from Pinlay. 

MACKINNON, G. M'Fhionghuin, M. G. Fionghuine, g. (M'V.), in 
Macfingon (1400), 0. G. Finguni, gen. (B. of Deer), Ir. 
Finghin, M. Ir. Finghin, Finnguine, E. Ir. Finguine : * Vimh- 
gonio-s, " fair-born " (fionn and gin) ; cf. for force and partial 
root Gr. KaWiyevr/s, and -yovos in proper names^ 

Mackintosh, G. Mac-an-toisich, the Thane's son (see tbiseach), 
M. G. Clanna-an-t6isaigh, Clans Mackintosh (M'V.), Toissick 
(D. of L.), Mackintoshes, Clann-an-toisigh (1467 MS.), early- 
documents M'-Toschy (1382). 

Mackiedy, G. M'Urardaigh, M'Urarthie, 1632 ; M'Quiritei, 
1626 ; Makmurrarty, 1547 ; Makwerarty, 1517 ; common in 
Bute and Arran of old, from Muircheartach, " sea-director " 
(muir and ceart) ; whence also M'Murfrie, M'Ufatrie. 

Mac-lachlan, G. M'Lachlainn ; see Lac/dan. 

Maclagan, G. M'Lagain (Lathagain in its native district of 
Strathtay), documentary Maklaagan (1525) : *M'Gillaagan, 
sed quid? 

Mao-laben, G. M'Labhruinn ; see Lawrence. 

Mao-larty, G. M'Labhartalgh and Lathartaich, from Flaith- 
hheartach, Eng. Flaherty : " dominion-bearing " or " princely- 
bearing " {s&ejlath and beartach). 

Mac-lean, G. M'lUeathain, for Grill' Sheathain, John or Seathan's 
servant, M. G. Giolla-edin (M'V.), Gilkeoin (1467 MS), 
documents Makgilleon (1390) ; from gille and Seathain (Iain) 
or Edin, John, the latter being the classic G. for the name. 
John means in Hebrew " the Lord graciously gave." 

Mac-lbarnan, so G. ; from GUI' Ernan, St Ernan's gille. The 
Latin name of this saint is Ferreolus, " Iron-one " ; from 

Mac-leay, G. M'An-16igh, or earlier M'An-l^ibh, documents 
M^Conleif (1498 in Easter Eoss), Dunslephe, gen. (1306-9, 
Kintyre), Dunsiaf Makcorry (1505), M. G. Duinsleibe, gen., 
Ir. BonnsUibhe, E. Ir. Duindslebe, gen. : " Brown of the Hill," 
from donn and sliabh (not ''Lord of the Hill," as other 
similaT names exist in dubh, e.g. DubhsUibhe ; see Mac-phee). 
Capt. Thomas regarded the M'Leays of the north-west as 
descended from Ferchar Leche, F. the physician, who gets 
lands in Assynt in 1386, being thus M'An-leigh, physician's 
son, Manx Cleg, Legge. The Appin M'Lea clan Englished 
their name as Livingstone, of whom was the celebrated 


Mac-lellan, G. M'Gillfhaolain, M. G. M'Gilklan (D. of L.), Gilla- 
faelan (1467 MS.), St Fillan's slave, E. Ir. Faeldn, 0. Ir. 
Fdilan, horn fdil, now faol, wolf, q.v. Hence Gilflllan. 

Mac-lbnnan, G. M'lllinnein, Servant of St Finnan, Ir. Mae- 
Gilla-finnen (common in fourteenth and fifteenth century), 
M. Ir. Finden, E. Ir. Finnian, Adamnan's Vinnianus = Finnio, 
Finnionis = Findbarrus ; from fitin, fionn, white : the full 
name, of which Finnan is a pet form, was Findbarr or "Fair- 
head," Eng. Fairfax. Skene deduced Mac-lennan from M. G. 
M^Gilla-agamnan, Adamnan's gille, documents Gilleganan 
Macneill (154.5), Gilleownan (1427). 

Mac-ledd, G. M'Ledid, M. G. M'Cloyd (D. of L.), M'Leod (MS. 
1540), documents Madoyde (fourteenth century), 0. G. Liot 
(B. of Deer), Norse Sagas Lj6tr, earl of Orkney in tenth 
century, and otherwise a common Norse name ; the word is 
an adj. meaning "ugly" (!), Got. liiita, dissembler, Eng. little. 

Mac-mahon, G. M'Mhathain ; see Afatheson. 

Mac-martin, G. M'Mhairtinn, no doubt for earlier Giilamartain, 
gen. (1467 MS., an ancestor of the Cameron chiefs) : Eng. 
Martin, from Lat. Martinus, the name of the famous fourth 
century Gaulish saint ; it means "martial." 

Mac-master, G. M'Mhaighistir, son of the Master. 

Mac-michael, G. M'Mhicheil, doubtless for earlier Gillamichol '; 
see Carmichael. 

Mac-millan, G. M'Mhaolain, M'Ghille-mhaoil, son of the Bald 
gille {at. M'Glaskan). To Maolan must be compared the 
Ogmic Mailagni. 

Mac-nab, G. M' An-aba, M. G. m'ynnab (D. of L.), M' An Aba 
(1467 MS.) : "son of the Abbot" ; see aba. 

Mac-nair, G. M'An-uidhir ; for Mac Iain uidhir, son of dun 
(odhar) John (cf. MaTtaneroy, 1556, now Mac-inroy, and 
Makanedvy, 1526, now Mac-indoe). Such is the source of the 
Gairlooh branch of the name. The Perthshire sept appears 
in documents as U'Inayr (1468), Macnayr (1390), which is 
explained as i/' An-oighre, son of the heir. if^Nuirs in 
Gowal (1685), John Makneivar (1546, in Dunoon) ; Tho. 
M'Nuyer (1681, Inverness). Prof. Mackinnon suggested 
M'An-fhuibhir, son of the smith or faber ; nor should 
M'An-jhuidhir, the stranger's son, be overlooked as a possible 

Mac-naughton, G. lil'Neachdainn, M. G. M'JVeachtain (1467), 
0. G. Nectan, Pictish Naiton (Bede), from necht, pure, root 
nig of nigh, wash. 


Mac-nee, G. M'Righ ; D. of L. M'onee, APNie, 1613; M'Knie, 
1594; M'Kne, 1480 (Menteith and Breadalbane). From 
mac-nia, champion ? 

Mac-neill, G. M'Neill, documents Makneill (1427). See ]!feiL 

Mac-nicol, G. M'Neacail, M. G. MNicail, from Lat. Nicolas, Gr. 
NiKoAas, " conquering people." Hence Nicholson. 

Mac-nish, G. M'Neis ; from M'Naois, the Naois being a dialectic 
form of Aonghus or Angus. 

Mac-niven, G. M'Ghille-naoimh, the saintly gille (cf. for form in 
Eng. Mae-glashan). Documentary form Gilnew M'llwedy 
(1506). The M. G. and Ir. Gilla Nanaemh, servant of the 
saints (1467 MS.), is a different name. The Ir. M^Nevin is 
for M'Cnaimhin. Mac Nimhein {Oranaiche 520). 

Macphail, G. M'Phail ; son of Paul. See Paul. 

Mac-phbe, G. M'a-Phi, M. G. M'a ffeith (D. of L.), M'Duibsithi 
(1467), documents Macdtiffie (1463), for Dub-shithe, Black of 
peace {dubh and slth). 

Mac-pheeson, G. M'Phearsain, son of the Parson, M. G. 
M'a pharsone (D. of L.), documents M'Inpkersonis (1594 Acts 
of Pari.). Bean Makimpersone (1490, Cawdor Papers), Mak- 
farson (l481, Kilravock Papers), Archibald M'Walter vie 
Doncho vie Persoun (who in 1589 has lands in Glassary 
of Argyle); Tormot M'Farsane (vicar of Snizort, 1526). The 
Badenoch M'Phersons are known as Glann Mhuirich ; the 
Skye sept are called Cananaich (from Lat. canonicus, canon). 

Mac-quarrib, G. M'Guaire, M. G. Gvuire, M'Guaire (1467 MS.), 
Macquharry (1481), M'Goire of Ulva (1463, Makquhory in 
1473) ; from Gadelic Guaire, *Gaurio-s, E. Ir. guaire, noble; 
Gr. yaC/jos, proud, exulting ; further Lat. gaudeo, rejoice, 
Eng. joy. 

Mac-qtjeen, G. M'Cuinn, documents Sween M'Queen (1609, Clan 
Chattan Bond), M'Qiieyn (1543, Swyne then also as a personal 
name, in Huntly's Bond), Makquean (1502, personal name 
Soyne also appears), M. G. Suibne, gen. (1467 MS., Mackin- 
tosh genealogy), M'Soenith (D. of L.), documents Syffyn 
(1269, the Kintyre Sweens), Ir. Suibhne (Sweeney), E. Ir. 
Svhne, Adamnan's Suibnevs : *Subnio-s, root ben, go: "Good 
going?" The opposite Duibne (O'Duinn, etc.) appears in 
Ogam as Dowinias (gen.). Cf. dubhach, subhach. Usually 
Mac-queen is referred to Norse Eng. Sweyn, Norse Sveinn, 
which gives G. M'Suain, now Mac-Swan, a Skye name. 
Pronounced in Arg. Mac Cui'ne or Cuibhne, for M'Shuibhne, 
which is the best spelling for Argyle. 

Mac-rak, G. M'Rath, M. G. gen. Mecraith, documents M'Grath 
(1383 in Rothiemurchus), Ir. Macraith (years 448, onwards) : 


'' Son of Grace or Luck," from rath, q.v. A personal name 
like Macbeth. 
Mac-eaild ; see under Harold. 
Mac-eanald, G. M'Raonuill ; see Ranald. 

Mac-eoey, Mac-euey ; see Rory. Documents give Mahreury in 

Mac-taggaet, G. M'An-t-Sagairt, son of the priest. 

Mac-tavish, G. M'Thiimhs, for M'Thd,mhais, son of Thomas or 
Tammas, M. G. Clyne Tawssi (D. of L.), documents M'Cawis 
and M'Cause (1494, 1488, in Killin of Lochtay). 

Mao-vioae, G. M'Bhiocair, documents Makvicar (1561, when 
lands are given near Inveraray to him) : " Son of the Vicar." 

Mac-vtjeich, G. M'Mhuirich, M. G. Mhuireadhaigh (M'V.) i the 
Bardic family of M'Vurich claimed descent from the poet 
Muireach Albanach (circ. 1200 a.d.). They now call them- 
selves Macphersons by confusion with the Badenoch Clann 

Neil, G. Niall, so Ir., E. Ir. Niall, Adamnan's Nellis, gen. : 
*N'eillo-s, *Neid-s-lo- ; see niata for root, the meaning being 
"champion." Hence JMac-neill. The word was borrowed 
into Norse as Njdll, Njal, and thence borrowed into Eng., 
where it appears in Domesday Bk. as Nigel, a learned spelling 
of Neil, whence Nelson, etc. 

Nicholson, G. M'Neacail ; see Mac-nicol. 

NoEMAN, G. Tormoid, T6rmod (Dial. Tormailt, for earlier Tor- 
mond), documents Tormode (David II.'s reign) ; from Norse 
Th6rm6Sr, the wrath of Thor, Eng. mood. The form 
Tormund alternates with Tormod (1584, 1560): "Thor's 
protection ; " whence the . Dial. Tormailt (cf. iarmailt for 
phonetics). Cf. Gearmailt, Germany. 

Pateick, G. P^druig, Paruig (with pet form Para), for Gille- 
phadruig, M. G. Gillapadruig, Ir. Pddraig, Giollaphdtraicc, 
0. Ir. Patrice ; from Lat. Patricius, patrician. Hence Mac- 
phatrich, Paterson. 

Paul, G. P61 (Classic), Fkl (C.S.) ; from Lat, Paulus, from paulus, 
little, 'Kng. few. 

Petee, G. Peadair ; from Lat. Petrus, from Gr. Hct/oos, rock, stone. 

Philip, so G. ; see Mackillop. 

Eanald, G. RaonuU, M. G. Raghnall (M'V.), Ragnall, Raghnall 
(1467 MS.), Ir. i?a(5TOa^/ (common); from Norse Rognvaldr, 
ruler of (from) the gods, or ruler of counsel, from rogn, regin, 
the gods, Got. ragin, opinion, rule ; whence Reginald, Rey- 
nold, etc. Hence M'B.aonuill, Mac-ranald, Clauranald. 

RoBEET, Raibert, Robart, Rob, M. G. Robert (D. of L.), Roibert 
(1467 MS.) ; from Eng. Robert, Ag. S. Robert, from hro, hrScf, 


fame, praise, and herht, bright, now bright, " bright fame." 
Hence Robertsons ( = Clann Donnchaidh), Mac-robbie. 

EoDERicK, RoRY, G. Ruairidh, M. G. Buaidri (1467 MS.), 0. G. 
£iladri, Ir. Ruaidhri, gen. Rtmdraeh (Annals at 779, 814), 
0. Ir. Ruadri, E. W. Rotri, Rodri ; from riiadh, red, and the 
root of r\gh, king? The Teutonic Roderick means "Famed- 
ruler" (from hr6^ and rilt, the same root as G. r\gK). The 
terminal -ri, -rech (old gen.) is a reduced form of righ, king 
(Zimmer, who, however, regards Ruadri as from N. Hrdrehr, 
but this in Galloway actually gives Rerih, M'Rerik, M'Grerik, 
1490, 1579, thus disproving Zimmer's view). M'Cririck still 

Ross, G. Rosach, Ros ; from the County name Ross, so named 
from ros, promontory. 

Roy, G. Ruadh, red. Hence Mac-inroy, earlier Makaneroy (1555), 
for M'lain Ruaidh, Red John's son. 

Samuel, G. Samuel, Somhairle. The latter r.2ally is Somerled, 
M. G. Somuirle (M'V.), Somairli (1467 MS.) ; from Norse 
-SumarlitSi, which means a mariner, viking, "summer sailor," 
from sumar and liSi, a follower, sailor. 

Shaw, G. Seaghdh, Englished as Seth ; evidently formerly Si'ach 
or Se'ach, Schiach M^Keich, Weem in 1637 ( = Shaw M'Shaw), 
Jo. Scheach, Inverness in 1451, Jo. and I'ho. Scheoch, king's 
"cursors" 1455-1462, Sytliack Macmallon in Badenoch in 
1224-33, Ferchar filius Seth there in 1234, M'Sithig in B. of 
Deer : *Sithech, M. Ir. sidkach, wolf, The female name 
Sitheag was common in the Highlands in the 17th century 
(Shiak, Shihag). The Southern Shaws — of Ayrshire and 
Greenock — are from De Schaw (1296), from Sc. and Eng. 
shaw, shaws ; the southern name influenced the northern in 
spelling and pronunciation. In Argyle, the Shaws are called 
Clann 'Mh\.G-g\a\\&-Sheathanaich. 

Simon, G. Sim. This is the Lovat personal name ; hence 
M'Shimidh, Simmie's son, the name by which the Lovat 
family is patronymically known. Hence in Eng. Sime, Mac- 
kimnlie, M'Kim, Simpson, etc. 

Somerled ; see Samuel. 

Sutherland, G. Suthurlanach ; from the county name. 

Taggart ; see M ac-taggart. 

Thomas, G. Tbmas, Timhus (M'F.), M. G. Tamas (1467 MS.). 
Hence Mac-tavish, Mac-combie. 

Torquil, G. Torcull (Torcall) ; from Norse Thorkell, a shorter 
form of Thorketill, which see under Mac-corquodale. . 

Whytb, G. M'lllebh^iu ; son of, the fair gille. See Bain above. 


f— 'r" 

William, G. Uilleam, M. G. William (1467 MS.) ; the G. is bor- 
--1 rowed from the Eng., 0. Eng. Willelm, Gar. Wilhelm, " helmet 
, _' of resolution" (from will and helm). Hence Mac-william. 


Beathag, Sophia, M. G. Bethog (M'V.), Bethoc {Chronicles of Picts 
and Scots : name of King Duncan's mother), for *Bethdc, the 
fern, form of Beathan, discussed under Mac-hean. 

Bride, Bridget, E. Ir., 0. Ir. Brigit, g. Brigte or Brigtae : *Brgnti 
(Stokes), an old Gaelic goddess of poetry, etc. (Corm.) ; 
usually referred to tne root brg, high, Celtic Brigantes, high 
or noble people ; Skr. hrhati, high (fem.) ; further Ger. herg, 
hill, Eng. hurgh. The Norse god of poetry was Bragi, whose 
name may be allied to that of Brigit. The name of the Gr. 
goddess AcjipoSiTT) (Bhrg-Ud) and the Teutonic name Berhta 
(from the same stem as Eng. bright), have been compared to 
that of Bridget (Hoffman, Bez. Beit, i*, 290) ; but this deriva- 
tion of Aphrodite ("foam-sprung"?) is unusual. 

Biorbh^il, Diorbhorguil, Dorothy, M. G, Derbhfdil (M'V.), Ir. 
Dearbhail, Dearbhforghaill, respectively translated by O'Dou- 
ovan "true request" (see aill) and "true oath" (E. \x. forgall, 
0. Ir. forcell, testimony, from geall). Hence the historic 
name Devorgilla. 

Pionnaghal, Flora, M. G. Fionnghuala (1469 MS.), documents 
Finvola (1463), Fynvola (1409), Ir. Finnghuala: "Fair- 
shouldered " ; from fionn and guala. 

M6r, M6rag, Sarah, M. G. M6r (M'V.), Ir. M(>r (year 916); from 
mdr, great, while Hebrew Sarah means " queen." 

Muireall, Marion, Muriel, Ir. Muirgheal (year 852) : Morirgeld, 
" sea-white " ; from muir and geal. 

Oighrig, Elghrig, Euphemia, M^ G. Effric (D. of L.), med. 
documents Africa, Ir. Aithbhric, older Affraic (two abbesses 
of Kildare so called in 738 and 833) ; from Africa 1 

Eaonaild, Raonaid, Rachel; from Norse RagnUldis, "God's 
fight." Of. Ronald. 

Sorcha, Clara, Ir. Sorcha ; from the adj. sorcha, bright, the 
opposite of dorcha, q,v. 

Una, Winifred, Winny, Ir. Una ; usually explained as from una 
(niina, M. Ir. =gorta), hunger, famine, whence the Ir. 
proverb : " Ni bhion an teach a mblon Una la na leath gan 
niina " — The house where Una is is never a day or half one 
without hunger." W. newyn. Cor. naun, Br. naon, M. Br. 
nafin, *novengo-, Eng. need. Cf. E. Ir. uinchi, scarcity, Eng. 
want, wane. Una, daughter of the King of Lochlan, is repre- 
sented by Keating as Conn Cedcathach's mother (second 



Ptolemy, the famous Alexandrian geographer, flourished in 
the second quarter of the second century. Nothing much ia 
known of his personal history, but his works on astronomy and 
geography dominated the world of learning and research for 
a dozen subsequent centuries. Ptolemy systematised the 
results of ancient research in these two subjects, adding some 
clarifying theories and facts of his own. That the earth was 
a sphere was a fact accepted by the ancient world of science 
ever since the time of Aristotle, but Ptolemy was the first to 
produce a rational plan for projecting the sphere, either in 
whole or in part, upon a plane surface. He is in fact our first 
real scientific map maker. 

Ptolemy's work on Geography is very properly entitled 
' ' Instruction in Map-drawing ' ' ; for, of its eight books, the 
second to the seventh merely contain tables of names of places^ 
with longitude and latitude attached, so as to be transferred- 
to the map. The first book gives instructions how to make 
the map with the proper projection. Ptolemy also drew 27 
maps himself, and maps that purport to be their descendantis 
are still found in the Ptolemy MSS. It is easy, however, to 
see that the real value of the work lies in the tables and not 
in the maps, whose accuracy, unchecked by the tables, could 
never, after so many centuries of copying, be depended on. 

Ptolemy's degrees of longitude start from the then known 
westernmost point of the world — the Canary Islands ; his. 
latitude, of course, begins from the Equator. His degree of 
latitude was estimated at 500 stadia, which is one-sixth too 
small ; his longitude degrees properly enough converge as he 
moves northward. His northernmost point of all is Thule, 
which he places in latitude 63 degrees. In regard to Britain 
his latitude on the south coast of England is 2 degrees too 
high, and by the time Scotland is reached this error is 
doubled ; the Sol way Pirth is put down as 58 degrees 45 
minutes, whereas it is 4 degrees less than this really. 

30 Ptolemy's geography of Scotland. 

While Ptolemy's outlines of England and Ireland are in a, 
general way fairly accurate, the fact that he places Scotland 
at right angles to England gives his map of the British Isles a 
grotesquely inaccurate appearance. On closer inspection, 
however, it will be found, when once this initial error is 
allowed for, that his outline of Scotland is as good as those for 
the two sister countries. Up to the Tyne and Solway, 
Ptolemy's map is as accurate as could be expected from his 
general work; but here, instead of continuing Scotland 
straight to the North, he turns it eastward, exactly 90 degrees 
Wrong. Many explanations have been suggested for this 
error ; the most satisfactory is that of Mr Bradley, who thinks 
that Ptolemy or a predecessor had England, Scotland, and 
Ireland first on three separate maps, and in fitting them 
together, he had placed Ireland too far North, and so, per- 
force, was led to place Scotland at right angles to England. 
In any case, latitude and longitude have shifted places as far 
as Scotland is concerned, and the Mull of Galloway is the 
furthest north point of Scotland according to Ptolemy. 

Of course the text is often corrupt, different MSS. pre- 
senting different readings. The latest and best edition is that 
of Miiller (Paris, 1883) ; he has collated some forty MSS., and 
he gives in his notes all the various readings, noting the MSS. 
in which they occur. I have followed Miiller's text in the 
translation and transcription of Ptolemy's Geography of 
Scotland, which I here present. The map which accompanies 
this paper has been kindly prepared by Mr James Eraser, 
C.E., Inverness. He has given the latitudes and longitudes 
of Miiller's text, but on these he has grafted the more or less 
rounded contours of the Latin Ptolemy map of 1478. This 
preserves the map from the odd look which such angular repro- 
ductions as those in Captain Thomas's maps always present, 
while accuracy, it is hoped, is not a whit sacrificed. 

After describing the " Britannic Isle of Ivernia," that ia, 
Ireland, Ptolemy sets about describing the situation of the 
Hebrides, which he places to the north of Ireland. Ptolemy's 
text runs thus : — 

There lie above Ivernia islands which are called ^budas, 
five in number, the westernmost of which is called — 

Longitude. Latitude. 

Degs. Mins Degs. Mins. 

^biida 15 62 

Ptolemy's geography of Scotland. 31 

The next to it towards the east is likewise 

Longitude. Latitude. 

Begs. Mins. Degs. Mins. 

^btida 15 40 62 

then Rhicina 17 62 

thenMalseus 17 30 62 30 

then Epidium 18 30 62 

And towards the east from Ivernia are these islands : — 

Monaoeda 17 40 61 30 

Mona island 15 57 40 

Adru, a desert island 15 59 30 

Limnu, a desert islajid 15 59 


The description of the north side, above which is the Ocean 
called Dvecaledonius : — 

Peninsula of the Novantse and 

Cape of the same name... 21 61 40 

Rerigonius Bay 20 30 60 50 

Vindogara Bay 21 20 60 30 

Estuary of Clota 22 15 59 20 

Lemannonius Gulf 24 60 

Cape Epidium 23 60 40 

Mouth of the river Longus ... 24 30 60 40 

Mouth of the river Itys [Eitis] 27 60 40 

Volas [or Volsas] Bay .: 29 60 30 

Mouth of the river Nabarus... 30 60 30 

Tarvedum or Orcas Cape 31 20 60 15 

The description of the west side, to which are adjacent 
both the Ivernic Ocean and the Vergionius Ocean. After the 
Cape of the Novantse : — 

Mouth of the river Abravannus 19 20 61 

Estuary of lena 19 60 30 

Mouth of the river Deva 18 60 

Mouth of the river Novius .. . 18 20 59 30 

Estuary of Ituna 18 30 58 45 

The description of the next sides looking south-east, to 
which is adjacent the Germanic Ocean. After Cape Tarvedum 
or Orcas, which has been mentioned : — 

Cape Virvedrum 31 ^^ 

CapeVerubium 30 30 59 40 

Mouth of the river Ila 30 59 40 

Si2 Ptolemy's geogeaphy op Scotland. 

Longitude. Latitude. 

Degs. Mine. Dees. Mins. 

High Bank 29 59 40 

Estuary of Varar 27 59 40 

Mouth of the river Loxa 27 30 59 40 

Estuary of Tvesis 27 59 

Mouth of th© river Caelis 27 58 45 

Capeof the Tsezali [Tasxali]... 27 30 58 30 

Mouth of the river Deva 26 58 30 

Estuary of Tava 25 58 .50 

Mouth of the river Tina 24 58 30 

Estuary of Boderia 22 30 59 

Mouth of the river Alaunus... 21 20 58 30 

Mouth of the river Vedra 20 10 58 30 

Th© Novantee dwell along the north side below the 
Peninsula of like name, among whom are these towns : — 

Lucopibia 19 60 20 

Rerigonium 20 10 60 40 

Below them are the Selgovse, among whom are these 
towns : — 

Carbantorigum 19 59 30 

Uxellum 18 30 59 20 

Corda 20 59 40 

Trimontium 19 59 

From these towards th© east, but more northerly, are the 
Damnonii, amiong whom are these towns : — 

Colanica 20 45 59 10 

Vandogara 21 20 60 

Coria 21 30 59 20 

Alauna 22 45 59 50 

Lindum 23 59 30 

Victoria 23 30 59 

More southerly are the Otalini [better Otadini], among 
whom are these towns : — 

Coria 20 10 59 

Alauna 23 58 40 

Bremenium 21 58 45 

After the Damnonii towards th© east, but more northerly, 
from Cap© Epidium about eastwards are the Epidii, after 
whom (the Cerones, then more easterly) the Creones, then the 
Carnonacse, then the Ceerem, and, most easterly and furthest, 
th© Cornavii. From th© Lemannoriius Gulf as far as the 


Estuary of Varar are the Caledonii and above them the Cale- 
donian Forest; from them more easterly are the Decantee, 
touching whom are the Lugi, and above the Lugi are the 
Smertse. Below the Caledonii are the Vaoomagi, among 
whom are these towns : — 

Longitude. Latitude. 

Degs. Mins. Degs. Mine. 

Bannatia 24 59 30 

Tamia 25 59 20 

Alata Castra (Winged Camp) 27 15 59 20 

Tvesis 26 45 29 10 

Below these but more westerly are the Venicones, among 
whom is this town : — 

Orrea 24 58 45 

More easterly are the Tsezali [Taexali] and their city : — 

Devana 26 59 

Islands lie adjacent to the Isle of Alvion at Cape Orcas : — 

Scetis Isle 32 40 60 45 

Dumna Isle 30 61 20 

Above which are the Orcades Isles, about 30 in number, 
the middle of which is 30 61 40 

And still further up than these is the Isle of Thtile, the 
parts of which lie — 

The westernmost 29 63 

The easternmost 31 40 63 

The northernmost 30 20 63 15 

The southernmost 30 20 62 40 

The middle 30 20 63 

I will examine the above names with a double purpose : 
first, to see if they, or the places they refer to, can be traced to 
modern times ; second, to discover, if possible, what language 
or languages the names belonged to. This last point practi- 
cally means that I am to discuss the Pictish question from a 
linguistic standpoint. Professor Rhys, as is well known, 
maintains that the Picts were non-Celtic and non-Aryan, a 
view which he has lately expounded afresh in an extraordinary 
paper in the Proceedings of the Society of Antiquaries, on the 
language of the Northern Picts. We shall see whether 
Ptolemy's names for the ancient Highlands and Isles involve 
necessarily non- Aryan or non-Celtic elements. Of the above 

34 Ptolemy's geoghaphy of Scotland. 

names, fortunately about two-thirds of them belong to the 
region of the Northern Picts. 

Ptolemy, as we know from Marcian, called the British Isles 
the Prettanic Islands, but the MSS. now have the more or 
less Roman form of Brettanic. Prettania is the real old Greek 
name of our Isles, and it is clear that the Roman Brittania is 
but a oorniption of it. Professor Rhys and others maintain 
that the two names are separate ; he says that Britain got its 
Roman name from the South of England tribes, who called 
themselves Brittones. But there is no authority for this. In 
fact, the name Prettania or Pretania has been preserved in its 
Brittonic form in the Welsh Prydain for Britain, and in 
Prydyn, the Welsh for a Pict = Gaelic Cruithne. Gaelic c 
answers often to Welsh p, and consequently Gaelic Cruithne, 
pre-historic Qrt-an-ic, is the same as the ancient Pretania ; in 
short, the Picts gave their name to the British Isles. Probably 
they were the only Celtic inhabitants of Great Britain when 
the Greek voyager Pytheas (circ. 300 B.C.) visited these isles. 
The name is allied to Gaelic cruth, form, appearance; they 
may have been called the " figured " or " painted " men, 
as old writers insist they were so adorned. 

Ptolemy's Alvion appears in Pliny and others as Albion ; 
he means by it Great Britain ; but the Gaelic population of 
both islands have always restricted this name to Scotland — 
Alba, gen. Alban. We may compare the Latin Alba to it; 
the Latin adjective alhus signifies white. The underlying 
meaning is the very common and, in this case, appropriate one 
of " White-land." Ivernia, " Ireland," is the Latin 
Hibernia — a piece of folk etymologising, for Hibernia in Latin 
means " Winter-land " (hibernus, winter). Another old 
Greek form of the name is lerne, which is exceedingly near the 
modern Gaelic sound. Soome think that an initial p has been 
lost, and explain the name as Piverion, " Pat or Rich-soiled 
Land," Greek Pieiria. It has, however, to be remembered 
that several rivers (the Scottish Earns, Find-horn, etc.) have 
the same exact name ; consequently it is either the name of a 
goddess, or a name applicable to both " island " and " river." 
(Compare the Teutonic ey, isle, from a root allied to aqua). 
Not only does 'Hire, Ireland, appear in river names, but we 
have at least two other island names applicable to rivers and 
estuaries — Ilea or Islay and Ptolemy's Dumna (compare the 
Irish Inver-Domnan, etc., from the goddess Dumna). Pro- 
fessor Rhys, of course, refers Ivernia incontinently to a non- 
Celtic source, and we hear much of the pre-Celtic Ivernians. 


If Sootland be shifted back into its proper place, the 
iEbudse Isles will be to the west of it, and suit, in a general 
way, the Western Isles. Ptolemy has 5 ^Ebudse ; Pliny says 
there are 7 Acmodse (^modse) and 30 Hsebudes ; Mela speaks 
of 7 Hsemodse. The name now appears resuscitated as 
Hebrides, by the misreading of a MS. copyist. Ptolemy has 
two islands of the same name — ^buda or Ebuda ; it is usual, 
from their position, to equate them with Islay and Jura ; but 
probably Capt. Thomas was right in 'identifying them with 
the two Uists, North and South. The Norse name for Uist 
is Ivist, and the first syllable is not unlike the first part of 
Ebuda.l Ehicina, Pliny's Ricina, appears in a few MSS. as 
Engaricina, and some have consequently been tempted to refer 
the name to Egg (Greek Enga may be Egga) ; but it seems 
certainly intended for Rathlin isle — Irish Reachrainn. 
Malseos is, of course. Mull; Adamnan calls it Malea. Dr 
Stokes equates the root mal with Albanian mat', height, 
border ; Lettic, mala, border. The change of a into u in 
modern times is caused by the influence of the e or i sound in 
the secooid syllable. The Norse name was Myl. The isle of 
Epidium has been equated by Skene with Lismore, and by 
Captain Thomas with Islay. As the name cannot be separated 
from Cape Epidium (Kintyre), Mr Bradley thinks that it is 
a bit of the Mull of Kintyre which was inserted on the Irish 
map which Ptolemy worked from — one of the three which he 
fitted so ill together. We shall treat of the name Epidium 
afterwards. It is generally agreed that Monaoeda, Pliny's 
Monapia, is the Isle of Man (Welsh Manaw), and that Mona 
is Anglesey. 

Ptolemy calls the Sol way Firth the Estuary of the Ituua ; 
this name is identified with that of the Eden river. Going 
westwards, or rather, according to himi, northwards, we first 
meet the river Novios, the Nith ; the word is the Celtic novios, 
new, Welsh newydd, Gaelic nuadh, and the word Nith is a 
Brittonic rendering of the old name. Next, in its proper 
order, we have the Deva or Dee ; the name simply means 
" goddess," and is one testimony, among many, of the worship 
of rivers and fountains, which Gildas (6th century) so bitterly 
complains of. There are many Deva's on Celtic soil both in 
Britain and Spain. Then comes the lena, but, unlike the 
Nith and Dee, the name is lost, and the guesses made vary 
between the rivers Cree (Skene) and Fleet (Thomas) and any- 

1 Since the above was written, Sir Herbert Maxwell suggests that 
Bute is the modern representative of ancient Bbouda. 

36 Ptolemy's geography of Scotland. 

way in Wigtoun Bay (Muller) . The Abravannus is identified 
with Luce river and bay (Skene and Bradley) and the Annan 
((Muller), the latter on account of the similarity of the names 
when Aher is removed from. Abravannus. It is usual to 
etymologise the name into Welsh Aher-afon, " river's mouth ' 
= Mouth oi Avon ; and this may be correct. Then we reach 
the Mull of Galloway, three times its normal distance away 
from the head of Solway (Ituna); under the name of the Cape 
of the NovantcB, the people who are represented as inhabiting 
the " chersonese " or peninsula which abuts here. The name 
has left no modern traces ; the root seems to be the same as 
that in Novios river — " New-comers?" 

Turning northward, or, according to Ptolemy, eastward, 
we have the Rerigonius Bay ; this is by general consent set 
down as Loch Eyan ; the form suits well enough the modern 
name. It might be divided into Re-rig-onios, " fore- 
stretching," the main root being the common one of reg, 
stretch, go. The Bay of Ayr is represented by Vindogara 
Sinus : there is no modern representative, but the vindo is the 
well-known Celtic adjective vindos, white, a nasalised form of 
the root vid, see. The -gara may be parallel to the common 
river-name of Garry — Gaelic Garaidh, being possibly from the 
root gar, sound, and meaning " brawling." The Clota is, of 
course, the Clyde; the Gaelic is Gluaidh, old genitive Gluade, 
Adamnan's Gloithe, Bede's (Welsh) Gluith; it is usiial to 
refer the word to the root Idu, klou, cleanse, the Latin duo, 
cleanse, cloaca, sewer. 

Next comes the Lemannonius Bay : we may take this form 
as the correct one, though many MSS. have Lelaanonius. By 
general consent the place meant is recognised as Lech Fyne ; 
Muller, Bradley, Thomas, and Stokes all agree on this. And 
it suits Ptolemy's position well enough, though Loch Long is 
technically more correct, where, indeed, Skene places it. 'The 
name still exists in that of Lennox, the older Levenax or 
Levanach, the Middle Gaelic of which is Leamhain. The root 
is lem, now leamh, an elm ; and we may compare the Helvetian 
Lake Lemannus. Some think that Loch Lomond is meant; 
at anyrate, they think it is its name that we have here got by 
some confusion or other. The Gaelic of Loch Lomond is Loch 
Loimean, but in old times it was called Loch Leven, a name 
which in Ptolemy's times would be Livona (Lei-vo-na, root lei, 
smooth, flowing, Greek leois i). The difficulty is not alto- 
gether with Lemannonius Bay, but with the Longus river, 

1 See further in lieliqui(e Celticce, vol. II., p. '551. 


which Ptolemy places next after Cape Epidium in a position 
that might suit, relatively to the other two places, the western 
mouth of the Crinan Canal. Here the river Add discharges 
itself into the sea; the name means the " Long River;" and 
hence Skene concluded that this Long river was Ptolemy's 
Longus. There are several objections to this theory. First, 
it takes for granted that the Gaels were the inhabitants of the 
district about the year 100 ; this may be true. Second, it is 
Ptolemy's practice to translate the native names into his own 
Greek, as witness High Bank and Winged Camp, not into 
Latin, as Longus would imply. Skene made the errot because 
he used a Latin miap and text, and he has even caused Mr 
Bradley to stumble after him. The name is the name of Loch 
Long, however much displaced ; it means ' ' Ship ' ' or 
" Navigable Loch," from Gaelic long, W Hong, ship. In 
fact, the Norsemen called this very firth Skipaf jordhr, that is. 
Ship Fjord. Dr Stokes thinks that Longus is a Celtic word 
cognate with Latin longus; but the word long or luing is a 
common name in the W'estem Isles, one or two islands going 
by more or less oblique forms of the name (Lunga, Luing, and 
two Longa's). The identification of Longus River with Loch 
Long implies much confusion on the part of Ptolemy or, 
rather, of his informants ; but when one looks at the numerous 
lochs and firths and headlands of the Clyde Firth and Argyll- 
shire Coast, one need not wonder that the Roman sailors 
blundered. It is right to say that Capt. Thomas identified the 
Longus river with Loch Linnhe — An Linn© Dhubh, or Black 
Linn. Cape Epidium (Mull of Kintyre), we shall discuss in 
the name of the Epidii. 

Starting from the Mull of Kintyre and ignoring Longus 
river, we next meet with Itis or Eitis river, which fits quite 
well as to distance from the Mull with Loch Etive. Tl^is 
identification has commended itself to Muller, Bradley, and, 
doubtfully, to Captain Thomas. The name suits well ; Etive 
in Modern Gaelic is Eitigh, in Middle Gaelic Eitchi (Story of 
Deirdre). The only difiiculty is that, if the t was single 
between two vowels, we ought now to have it aspirated. It 
has, however, to be remembered that Etive is doubtless a word 
borrowed into the Gaelic, and in that case the rule does not 
always hold (witness the early borrow sagart, from sacerdos). 
Stokes gives the root as ei or i, to go, as in Lat. itum, iter, etc. 
Some compare the Gaulish Portus Itius, whence Caesar started 
for Britain. Skene places Itis at Loch Carron. 

Accepting Loch Etive as Ptolemy's Itis, we find that his 
distance therefrom to the Volsas (Volas) Bay will bring us to 

38 Ptolemy's geogbaphy of Scotland. 

the neighbourhood of Loch Alsh, a name that wonderfully fits 
with that given by the old geographer. The root may be vel, 
vol, to " well," " roll;" German wolle, a wave. MuUer, 
Skene, and Thomas place the Volsas Bay at Loch Broom, a 
view that is tenable enough if the longitude of the Varar, 
Loxa, etc., on the east coast are considered. If Loch Alsh is 
the Volsas Bay, then the river Nabarus, which is undoubtedly 
the Naver, is much too far south — only one degree away from 
Loch Alsh. Any way we take it, there must be a discrepancy. 
As can be seen, Ptolemy ignores Cape Wrath, though many 
writers think that this is his Tarvedum Promontorium, notably 
Mr Bradley, who thinks that Ptolemy has misplaced the 
Naver ; in fact, he thinks that Tarvedum and Vervedrum 
should come before the Naver. But this is very unlikely, as 
we shall see. In many MSS. Nabarus is given as Nabseus, but 
there is no doubt in the mind of any one that the river is the 
Naver. The root seems to be nav, swim, etc., whence navis, 
a ship ; in short, the meaning of the word is much the same as 
we found in that of Loch Long. The Modern Gaelic is Nawir, 
the preservation, such as it is, of the v showing a borrowing 
from the previous Pictish tongue. 

After the Naver comes Cape Tarvedum or Orkas, which 
Captain Thomas and Mr Bradley identify with Cape Wrath. 
Mr Bradley rests his case on his derivation of Vervedrum, 
which he thinks is the progenitor of Farout in Farout' Head — - 
an impossible derivation. It is altogether a needless disloca- 
tion of Ptolemy's positions ; he means the three or four heads 
to the north and east of Caithness — Holburn Head, Dunnet 
Head, Duncansby Head, and Noss Head. Tarvedum is given 
by Marcian as Tarvedunum, that is. Bull's Dun or Fort; 
compare the Tarodunum of Gaul with like force. The meaning 
may, however, simply be Bull's Head. The point meant is 
either Holburn Head, near Thurso, or Dunnet Head, also 
forming an outpost to Thurso Bay. As a proof of our 
identification, Thurso itself is the Norse Thjorsa or Bull's 
Water ! With it may be compared the Icelandic Thjorsa or 
Bull's Water of modern times. Ptolemy gives the cape a 
secondary name — Orkas ; it seems to me that he means the two 
sentinel capes of Thurso Bay — Holburn and Dunnet Heads. 
Duncansby Head is called Virvedrutn Cape ; all writers aire 
agreed upon this, Mr Bradley excepted. He thinks that Far- 
out Head is meant ; he analyses Vir-vedrum into the preposi- 
tion ver, the Gaulish form of the Gaelic for or far, Lat. s-uper, 
iireeh uper ; it means " upon " or " exceeding." The vedrum 

Ptolemy's geography of Scotland. 39 

he equates with the Pictish fothar, appearing in Dunottar 
(Simon of Durham's Dum-foeder). The vir is doubtless the 
prep, ver; but vedrnm can hardly be fothar, for the latter 
word itself is simply a prefix word — a preposition, seemingly 
of like meaning with Gaelic for. The Wear river is called by 
Ptolemy Vedra; and Dr Stokes suggests a oooinection with 
O. Slavonic Vedru,^ clear. This would give a meaning in each 
case of Cape Clear and Clear River, which are, as to significa- 
tion, quite satisfactory. Cape VerCibium, or Noss Head, also 
contains the prep, vei-; the root ii,b has been happily referred 
by Stokes to the Irish word ubli, sword-point, doubtless allied 
to the English weapon (root veb, uh). This would give the 
meaning of the word Verubium as " Sword Head." 

Turning now southward, or westward according to Ptolemy, 
we come to the river Ha. By almost common consent this is 
regarded as the Helmsdale River, called in Gaelic Ilidh, Eng. 
IJlie, Sir Rob. Gordon's Vlly. The name Ha is common as a 
river name in Scotland (spelt Isla usually), and there is also 
the Island of Islay so named. In this we must remember the 
parallel case of Erinn in being used both for rivers and for the 
Island of Ireland. Stokes suggests a reference to the root in 
German eilen, to hasten, go. Skene draws attention to the 
fact that the syllable il enters largely into Basque topography. 
A degree (of longitude) further south is " High Bank," which 
Skene identifies with the hills north of the Dornoch Firth, but 
which most writers regard as the Ord of Caithness misplaced. 
It is likely the Ord of Caithness that is meant, and some seek 
the Ha north of it in the Latheron district, but without 
success. Berriedale Water may have also once been an Isla; 
witness the North and South Esks. The Varar Estuary is 
undoubtedly the Inverness and Beauly Firth. The name still 
exists in the River Farrar, and glen of Strath-farrar. The 
root is var, which may mean " winding," " bending;" com- 
pare Lat. varus, varius. We are now at the innermost corner 
of the Moray Firth ; and it may be remarked that Ptolemy has 
a wonderfully accurate account of this part, indeed of the 
whole, of the eastern coast of Scotland. 

The river Loxa is represented as in the same parallel (that 
is, longitude) as Varar, but half a degree to the north. That 
is how the best MSS. have it; other MSS. place the Loxa 
between High Bank and Varar, equating it with the Cromarty 
Firth (Captain Thomas), or the Loth, an insignificant river in 

1 Root vid, see, a3 in Gaelic fionn, white. But ved, wet, suits the 
phonetics better. 


Loth Parish (Bradley). The Loxa ooight naturally to b© the 
Nairn by position ; but the name is identified by Skene, Stokes, 
and others with that of the Lossie, far away from Ptolemy's 
place for it on the map. The phonetic diiSculty here is a 
racial one; from an early Pictish x, we should expect a later 
ch, that is, if the Pictish was a Brittonic language and treated 
X as the other Brittonic languages did. Compare Ochil of the 
Ochil Hills and the Welsh uchel, high, Gaelic uasal, Gaulish 
uxellos. The form lok-s may be from one or two roots, and 
may mean " oblique," " shining," etc. Measuring from 
Varar, we should put the Tvesis Estuary about Cullen ; it is 
doubtless the mouth of the Spey that is meant. On this all 
the authorities are agreed. The names seem also allied ; Spey, 
Gaelic Spe, may come from Spesi-s or Speisi-s, a Celtic sqvei, 
to vomit; Gaelic sgeith, Welsh chwyd, vomo ; compare for 
force the old Italian river Vomanus. Ptolemy's tv initial is an 
attempt to reproduce the initial Pictish sound which has now 
settled into the very non-Gaelic form of sp in Spey. Half-way 
between Spey and Kinnaird Head is the Caelis (Greek kailis) 
River, or Celnius, which suits the position of the important 
river Doveran or Deveron, but which in name fits Cullen and 
Cullen Water (Welsh makes original ai into u; hence Kailnios, 
which twioi MSS. give, represents admirably a later Cullen). 
Doveran is a Gaelic name and a late one ; as the earlier form 
DufEhern "shows, it means the Black Barn opposed to the 
Pindhorn or White Earn. The root kail is in modern Gaelic 
caol, narrow. Kinnaird's Head is called the Cape of the 
Tsezali or Taexali (Taixali), a name that should produce in 
later times a Pictish (British) Tuch-al or a Gaelic Taosal ; the 
parish of Tough in mid Aberdeenshire ideally represents the 
British form of the root. 

Turning southward, we come to the River Deva, now the 
Dee. Skene accepts the bad reading of one MS., which gives 
Liva or Leva, and identifies it with the North Esk. Next 
comes the Estuary of the Tava, the Tavaus of Tacitus, which 
in position suits the Esk, but in name and in reality means the 
Tay. The name Tava appears on Brittonic ground in the 
Devon Tavy and the Welsh Tawe, and there is a Welsh adjec- 
tive taw, signifying " quiet," " gentle," to which Gluck 
equates the Gaulish Tavia, Tavium, and the woman's name 
Tavena. Between the Tay and the Forth. Ptolemy places the 
river Tina or Tinna; by position, of course, it suits the Tay 
best. The river meant is the Eden, which makes a considerable 
bay near St Andrews. Many think that the Tyne, of New- 

Ptolemy's geography or Scotland. 41 

castle, is meant, but this is unlikely, because this portion of 
the coast was possibly the one best known to the Roman fleets, 
as we can easily guess from Agricola's campaign. If it be the 
Eden, then the remarks on the Eden from which we started, 
may apply to its derivation. Otherwise Tina or Tinna may 
be referred to the root ten, stretch, pull, Welsh tyn. 

The Estuary of Boderia is undoubtedly the Firth of Forth ; 
this name Tacitus gives as Bodotria. By combining the two 
readings we may arrive at a form Bodertia, the first portion of 
which may be the well-known Pictish form Pother, so common 
in place-names as a prefix, latterly dwindling into For (compare 
Fothuirtabhaicht, now Forteviot, Fordun from Fotherdun), 
or provected in Scotch to Fetter (Fettercairn, Fetteresso). It 
is possibly terminal in Dunottar, anciently Duin Foither, 
Oppidum Fother, Dun feeder (Simon of Durham for latter). 
Dr Stokes suggests a connection with Irish foithre, woods; but 
the Pictish fother points as likely to an older voter, a com- 
parative form of the prep, vo, under, and comparable to a 
Greek wpoteros. The old Picto-Celtic form of Forth may have- 
been Vo-ter-tia, which with the hardening of the v and 'the 
softening of the t (to d), which were in pro'cess probably as 
early as the first century, would give us the Boderia. or Bodo- 
tria. of the Classical writers. Forth seems to be the descendant 
of the word which Bodotria stands for. The 12th century 
writer of " De Situ Albaniae " says the river is called Froch 
in Gaelic (Scottice) and Werid in. Welsh (Britannice) — Eng. 
Scottewatre, that is, Scottish Sea. In two Irish versions of a 
poem on the Picts, added to the Irish mediseval Nennius, we 
are told the Picts took Alba " O crich Cat co Foircu (or 
Foirchiu)," that is, " from the bounds of Caithness to Forth." 
Zeuss compares Bodotria to the form buadarthe, turbulentus, 
applied to a stream in an old Irish gloss, and no doubt a 
shorter form of the word huadar, that is, hodar, would do ; 
but then the modern name Forth, which seems connected with 
the Classical names, must receive a separate explanation. 

The next point on the coast noticed by Ptolemy is the 
mouth of the river Alaunus; the river named is the Alne of 
Northumberland, surely insignificant compared to the Tweed, 
which is ignored. Captain Thomas suggests that the Tweed 
is meant but the Alne named. In a similar way, the Tyne is 
passed unmentioned, while the Wear is taken, under the name 
of Vedra. For its derivation, see Cape Vervedrum. There 
was another Alaunus in the south of England, identified with 
the Axe, and two cities in France and two in Britain called! 

42 Ptolemy's geography of Scotland. 

Alauna. There are at least three Scottish rivers called Allan, 
and this is supposed to be the modern form of ancient Alaunos, 
or, the more Celtic, Alauna. There is a Welsh Alun river, 
..and the Cornish Camel is also known as the Alan. The word 
likely divides into Al-auna, and possibly the root is pal, as in 
Latin ■palus, marsh. 

Let us now consider the sixteen or seventeen tribes that 
Ptolemy divides Scotland among. The Novantae occupied 
Wigton ; w© have already regarded them as the " New-comers, ' 
root nov, new. Eastward to the head of the Solway lay the 
Selgovae, whose name still survives in Solway Firth ; Lhe rcot 
is selg, which in the Celtic tongue means " hunting;" the 
Selgovae were the " Huntsmen." The great tribe of tha 
Damnonii occupied the counties of Ayr,- Lanark, Renfrew, 
Dumbarton, Stirling, Menteith and Fothreve of the western 
portion of Fife (Skene). They are generally in name regarded 
the same as the Dumnonii of ancient Devon, to which they 
gave that name. Prof. Rhys calls the Damnonii a Brythonic 
people, and finds remains of their name in that of the river 
Devon in Perthshire. If the Damnonii are in name the same 
as the Dumnonii, the root is the very common Celtic one of 
Dumnos or Dubnos, " world," the modern Gaelic domhan, 
allied to Eng. deep. East of the Selgovas and Damnonii lay 
the Otadini (long o), along the east coast from the Wear to 
the Firth of Forth, if not into Fife ! So awkwardly does the 
town Alauna fit the position of the Estuary of Boderia that 
the situation of the town suits only the isle of Inchkeith. 
Otalini is the reading of five good MSS., Otadini or Otadeni 
that of 15, and Tadini or Gadeni that of some others. The 
■preferable reading is Otadini, which suits the old Welsh name 
of the Lothian district, viz., Guotodin. 

While the tribes we have just enumerated are said to have 
possessed towns, which are duly named and " positioned," a 
matter which helps the identification of the tribal localities, 
the next ten tribes are slumped together townless, and with 
little or no guidance as to their position. First come the 
Epidii, touching the Damnonii to the north-west, and starting 
from Cape Epidium, as Ptolemy says. We may assign them 
Kintyre and Lorn. The name is from the root epo-, the 
Gaulish for " horse," Gaelic each. The ancient Gaelic name 
would have been Eqidios ; indeed the name exists in Adamnan's 
personal name, Echodius, and the later Eachaidh, which in 
Gaelic passed into Eachuinn. Dr Stokes, however, thinks the 
:Toot is peku, cattle, Lat. fecu; he would give the oldest form 

Ptolemy's geography of Scotland. 43' 

as (p)ekvidioi, " cattle-holders," the later Irish personal name 
being Eochaid ; but the other derivation seems the right one, 
for, as a matter of fact, the root peku appears nowhere else in 
Celtic. Next are the Cerones or Creones, for the MSS. differ 
as to whether these were two or one people. The root of the 
name is either ker or kre; it is a root of several meanings, the 
chief one of which is to " cut, divide, throw." Varioua 
authorities see remains of the name in the West Coast loch 
names — such as Crinan, Creran, etc., also Carron, Keiarn, 
Kearon. Captain Thomas extends them from Crinan to Loch 
Leven, and finds the name in Creran. The Carnonacse, 
according to most writers, occupied Wester Ross, and, of 
course. Loch Carron has been equated with the name. The 
true derivation seems to be earn, " a hill," common to all the 
Celtic tongues; they were the men of the " Rough-bounds," 
or Garbh-chriochan — ^the " Cairn-men." The use of Carn or 
Cairn for mountain names is peculiar to Pictland and Wales. 
Dr Stokes connects the 8th century nam© Monith Carno, the 
scene of a battle between two rival Pictish kings, fought in 
729, near Loch Lochy (?). The Csereni, or, properly, Caireni,. 
may be placed in Western Sutherland up to near the Naver. 
The root cair is that of *cairax, the modern caora, sheep ;, 
compare the Caeroesi of Gaul. Animal names giving names to 
persons or nations is not an uncommon phenomenon. The 
Cornavii occupied Caithness, the " horn " or corn of Scotland. 
There were Cornavii between the Dee and the Mersey in 
England, and Cornwall still holds the name, standing for 
Corn-Wales, " the Welsh of the Horn." 

Neighbours to the Cornavii southwards were the Lugi, 
occupying easter Sutherland. Around Loch Shin were the 
Smertae, and Easter Ross was occupied, up to the Varar 
Estuary, by the Decantse. The root lug of the name Lugi 
appears in many Celtic names, both on the Continent and in 
Ireland. Indeed, there was a Lugi tribe in Mid-Germany 
contemporary with the Highland Lugi. The god of light and 
arts among the Gael was called Luga of the Long Arms ; and 
the old name of Lyons was Lugdunum, explained by an old 
glossary as " desiderate monte " — the desirable town. Dr 
Stokes refers the root lug to a Celtic base corresponding to 
Ger. loken, allure, Norse lokka. In that case the Norse god 
Loki is Aryan cousin, probably, to Luga, though the former 
is the god of evil enticement, while the Celtic Lug is alluring 
by good. The Smertos or Mertse also shows a common root; 
we have the personal names Smertalos (Cumberland inscrip- 

44 Ptolemy's geography of Scotland. 

tion), Smertulifcanos, Smertomara, Ad-smerios, etc. ; and the 
goddess Minerva of tlie Gauls, called Ro-smerta. These Dr 
Stokes refers to the root smer, to shine. The Decantse are 
paralleled by an ancient people of North Wales — the Decanti, 
or Decangi (?), now Degannwy. The name seems also to be 
found on the Ogam inscriptions as Deceti; there is also the 
Decetia of Cassar. Dr Stokes gives the root as dec, Latin decus, 
glory, Eng. <iecorous. 

The Caledonii or Caledonians are definitely located ; they 
stretched, s ys Ptolemy, from Lemann Bay to the Estuary of 
Varar — from Loch Fyne to the Inverness Firth ; above them, 
that is, west of them, is the Caledonios Drumos or Forest. 
This last, as Skene said, is Drum Alban and the western 
Grampians. A line from Loch Fyne to Inverness goes right 
along Drum Alban for half the way ; in fact, the district so 
defined is an impossible one for two reasons. The Caledonians 
were east of Drum Alban ; secondly, Dunkeld, which most 
writers allow as containing their name, shews that Perthshire 
was occupied by them. In fact, the Caledonians inhabitated 
Perthshire and easter Inverness. Much discussion has raged 
round the name, and authorities are by no means agreed yet. 
Professor Windisch gave as the root lof Caledonian the form 
Cald-onios ; the root cald in Gaelic and Welsh means " wood " 
— Gaelic coille, Welsh celli. Hence Caldonii or Caledonii 
meant " Woodlanders." As a further proof, the name Dun- 
keld is in old Gaelic Dun Calden, now Dun-Chaillinn ; and 
there is also the mountain Sidh-Chaillinn in mid Perthshire. 
The combined sound Id was separated by the Romans and a 
vowel e introduced ; this was mistakenly lengthened by 
Ptolemy and his imitators. The Welsh forms show Celidon, 
but are evidently founded on th^ Latin pronunciation of 
Caledonia. Dr Stokes separates Caledonii from both Dun 
Calden and from cald, coille, wood. He cannot agree that the 
root in Caledonia can be Calden and cald. Professor Rhys, on 
the other hand, allows that Dun Calden contains the name 
Caledonia, but he denies that either can be of Celtic origin, 
much less akin to the root cald. The Dve- Caledonian Sea or 
Western Ocean has a puzzling name ; the dve has been ex- 
plained as meaning " two." This postulates two Caledonias, 
and this there was, if Perthshire was their southern and 
Inverness-shire their northern seat, and according to all 
evidences the Caledonians were an inland people ; it is difficult 
to understand how they could have given their name to the 
Western Ocean, unless, indeed, the northern half occupied all 

Ptolemy's geography op Scotland. 45 

Inverness-shire, west as well as east. In the wars of Severus 
so important were the Caledonians become that only they alone 
are mentioned for the northern Highlands, the other tribe 
being the Maeatee, near Agricola's wall. 

The Vacomagi, according to Ptolemy, occupied territory 
east of the Caledonians and coterminous with theirs through- 
out. What suits his figures is the stretch of country which 
begins on the Moray Firth with Elgin and Banff, includes 
Western Aberdeenshire and Perthshire, or, at least, the 
eastern portion of it. One of their four towns — Bannatia — 
would fall about the Earn or Almond, near Crieff, while the 
northernmost town is identified, probably rightly, with Burg- 
head, and is called " Winged Camp." The name Vacomagi 
has been explained by Mr Bradley and Professor Rhys as 
" men of the open plains," in opposition to the Caledonian 
Woodlanders. The stretch of country allotted to them by 
Ptolemy, intersected as it is with the Grarapians and its south- 
tending spurs, would hardly gain them this name ; but possibly 
they really occupied Forfar, West Aberdeen onwards to the 
M'Oray Firth, leaving Perthshire to the Caledonians. As to 
the derivation, the form vacos is very common in Celtic names, 
both as prefix and affix, but, as Gluck remarked, its meaning 
is obscure. The Welsh gwdg, empty, which Mr Bradley, and 
Professofr Rhys's derivation brings forward as a parallel, is 
apparently borrowed from Latin; Dr Stokes, however, allows 
it as a native word in his Comparative Dictionary, just pub- 
lished. The form magi may be equated wifli old Gaelic mag, 
plain, now magh. No trace of such a name as Vacomagi now 
exists. The Taexali of Aberdeenshire we have already dis- 
cussed ; and Ptolemy's only other tribe is the Ven(n)icones, or 
Venicomes ; this name is found also, but only once or twice, as 
Vernicomes. They occupied Easter Fife, and perhaps 
stretched northward into Forfar; Ptolemy makes them and 
the Taexali fill the whole coast, apparently, from Kinnaird 
Head to Fife Ness. Modern editors prefer Vernicones as the 
name of this people ; Professor Rhys once explained the name 
as " Marsh-hounds," adducing the Welsh gwern, a swamp, 
and cwn, hounds, as complete parallels. It may be remarked 
that " dog " names were common and popular among the 
Gaelic Celts, and the derivation is, therefore, not to be rejected 
on any idea that such a name would be insulting. But verno 
may also mean " good " and " alder," and, then, there is the 
terminal comes as against cones in the MS. readings. Possibly 
Professor Rhys's derivation is the best one. If we take Ven 

46 Ptolemy's geography of Scotland. 

instead of Vern, then we have the well-known Celtic root for 
" kin," " friends," Gaelic fine, seen in names like Veneti, etc. 
Ptolemy mentions 21 towns as existent in Scotland in his 
time ; they are mostly in the south, the Highland tribes having 
none, we may say. What exactly Ptolemy meant by his 
" towns " it is hard to say, for of regular towns in a Roman 
sense there were none. Possibly defensive positions — the diina 
or fortified hills and the strongholds in woods or by rivers — are 
meant ; and certainly within the lines of Roman conquest and 
campaigning the " tawns " appear to have been so many 
Roman positions taken up and occupied by Roman caomips and 
forts : witness such a name as Victoria, the position of which 
is naturally assigned to some place in Perth or Fife that 
presents prominent Roman remains. Henoe we may account 
for so many towns being named by the rivers on which they 
were placed ; the town, as it were, was ' ' the Camp on the 
Allan " or whatever river it may have been. Again Ptolemy 
is very inaccurate in his account of the position of the towns, 
even in England. Says Mr Bradley: — " No reliance can be 
placed on any of Ptolemy's indications of the position of inland 
places not otherwise known to us ; and the limits of the tribal 
territories are dependent almost entirely on the situation of 
the towns. Under these circumstances, it does not seem that 
Ptolemy's internal geography of Britain is likely to repay the 
trouble of a minute examination." The case in Scotland is 
much worse than with England ; there certain names are recog- 
nisable, but in Scotland no town name has survived from 
Ptolemy's time. In these circumstances, we shall pass them 
in very brief review. 

The Novantse had two towns — Lucopibia and Rerigonium. 
The former is placed by Skene at Whithorn, once St Ninian's 
Candida Casa. The licco of Lucopibia means " white " in 
Greek and " shining " in the Brittonic languages ; so that the 
English, Latin, and Greek are all translations or adaptations 
(Greek) of the Celtic original. Others see the name preserved 
in Luce and Glen Luce, and transfer the town to the latter 
pleuje. Rerigonium was on Loch Ryan, doubtless, where 
Roman works can still be seen. The Selgovae had four towns 
• — ^one was Carbantorigon, possibly at the Moat of Urr, between 
the Nith and the Dee (Skene) ; Rhys thinks the name is a 
Celtic degradation of Carbantorion, " chariot town," for 
Garhanton meant " chariot." Uxellum or " High-town " 
(Welsh, uchel; Gaelic, uas, uasal) may have been Wardlaw 
Hill, at Caerlaverock (Skene) ; while Corda, whose derivation 

Ptolemy's geogeaphy of Scotland. ,47 

■ is doubtful, was possibly at or about Sanquhar. Trimontium, 
or " Three Mount," suits the Eildoij. Hills for meaning, but 
Skene, for position and for the Roman works there, places it at 

The Damnonii had six towns — the first town is Colania, 
near the sources of the Clyde, making a frooitier post on a 
northward march ; second, Coria, which Skene places at 
Carstairs, where numerous remains, both Roman and Natiy?, 
have been found; third, Vindogara, whose derivation we dis- 
cussed already, which may have been at Loudon Hill, in 
Ayrshire, where remains of a Roman camp exist ; fourth, 
Alauna, which Skene places at the junction of the Allan and 

■ Forth, and which would form a defence against a foe advancing 
into Caledonia. It was somewhere on the Allan ; Rhys places 
it at the famous Ardoch, near the Allan. The fifth town is 
Lindum, which Skene places at Ardoch; the word mealns 
" water, Zjn?i;" for name Lindum would suit Lin-lithgow best. 
The town of Victoria, which, in Celtic, would be Bioudica or 
Pictish Budic, must have been a Roman station, possibly at 
Lake Orr, in Wester Fife (Skene) where Roman remains exist. 

The Otadini had three towns: first, Coria, which is vari- 
ously placed at Peebles and at Carby Hill, Liddesdale (Skene) ; 
second, another Alauna, which, as was said, suits Inchkeith by 
position, but is possibly Alnwick misplaced. Bremenion, 
whose root hrem means " roaring," is placed by Skene at 
High-Rochester in Redesdale, where traces of the Romans still 

We are on the confines of the Highlands, if not actually 
in them, when w© come to the Yaconaagi. In any case they 
were a great Pictish tribe. They had four towns: first, 
Bannatia or Banatia, which is variously placed on the Earn at 
Strageath, the Almond at Buchanty (Skene, Rhys), or the 
Garry in Atholl (Thomas). The editor of the Monumenta 
Historica jiritannica. places it at Inverness and Bonq. The 
root ban or bann is found often on Celtic ground : it means 
" white," " milk," " bond," etc. The second town is Tamia, 
which Skene places on the isle of Inchtuthill in.the Tay,, where 
numerous remains exist; the root tarn is common, especially 
for river names, and possibly means " dark." The " Winged 
Camp " is by common consent allocated to Burghead; while 
Tvesis is placed by Skene at Boharm on the Spey, for it is but 
the river name used as a town name. 

The Vernicones have been allocated only one town — Orrea, 
which has be«n variously placed by modern writers— at the 


junction of the Orr and Leven in Fife (Rhys), at Abernethy 
(Skene), and at Forfar (Thomas). Somewhere on the Orr 
seems best. Skene has suggested that Orr, the river name, is 
connected with Basque Ur, water ; so, he thinks, are the several 
rivers of similar names which we have — Oure, Ure, Urie,i 
Orrin, and Ore. This should delight Prof. Rhys. The root 
or, however, is a good Aryan one, and signifies to " run," 
" start;" the Norse orr means swift; and the root por, which 
also in Celtic results in or, gives like meanings — " passing 
through," etc. The Tsexali are represented as having one 
town — Devana. Skene places this, mostly because of simi- 
larity of name, at Loch Daven, near Ballater. The phonetics 
are unsatisfactory in two ways ; the a does not correspond to e 
in Devana, and it is unlikely that v oould be preserved in such 
a unique word. In fact, the v has disappeared out of the 
name Deon, the Don now, from which Aberdeen has its namie. 
The Gaelic is Dian, for ancient Divona, or rather Deivona. 
This is undoubtedly Ptolemy's word as well, and possibly 
Aberdeen, or rather Old Aberdeen is meant. The nam© 
means " goddess," and is found in Gaul ; 2 for the idea under- 
lying it, see the remarks on the Dee or Deva. 

Ptolemy places four islands, or island groups, adjacent to- 
his Cape Orkas. The first is known in the best MSS. as 
Skitis or Sketis (once only), while most MSS. give the form. 
Ocitis. Ptolemy places Skitis about 70 miles N.E. of Cap» 
Orkas, and, owing to the form Ocitis, many writers consider it 
as having been one of the Orcades islands. It is, however, 
more probably the Isle of Skye m^isplaced, a view which com- 
mends itself to MuUer, Thomas, and Stokes. The latter says, 
that it is " the wing-shaped island of Skye; Norse Skidh,-^ 
Irish Scii (dat. case, date 700 in Annals of Ulster) ; Adamnan, 
Scia; gen., Sceth (date 667 in Annals of Ulster), Scith 
(Tigemach, 668); means wing, Ir. Sciath, Sciathan." Dr 
Stokes' derivation is the one usually accepted ; the Norse 
SMdh, which is possibly influenced by " folk-etymology,"" 
means a " log," " firewood," " tablet," and is allied to 
another Gaelic Sgiath, a shield. It is interesting to note that 
the Dean of Lismore refers to the island as ' ' Clar Skeith ' ' — 
the Board of Skith. 

1 The u in most of these oases is long. Inver-urie appears in 1300 as- 
Inver-vwry, though the modern spelling existed in 1199. The derivation' 
suggested is from iubhar, the yew, the G-aulish 'Eburo, which appears in 
BO many ancient names of places, rivers, and peoples. 

2 Ansonius (4th century) explains it thus: " Divona, fons addite divis.'' 

Ptolemy's geography of Scotland. 49 

The second island is called Dumna ; he places it some 60 
miles north, that is west ( ?), of the mouth of the Naver, and 
the Orcades Isles only twenty miles further off northwards. 
Pliny mentions Dumna, but places it along with Scandia or 
Scandinavia. Mr Elton thinks it is on© of the Orkneys, 
Captain Thomas says that it is usual to connect it with Stroma 
(•■' Current Isle "), and Mr Bradley considers it either to be 
Skye or the Long Island. It may be the Long Island; the 
namie seems to contain l^he root which w© have already dis- 
cussed in the case of the Damnonii or Dunmnonii, the u of 
which is also long.i 

There are thirty Orcades Islands, says Ptolemy; other 
writers, such as Pliny, Mela, and Solinus, mention them and 
their numbers (40 or 30), and Tacitus tells us that Agricola's 
fleet subdued them. The nam© is still with us in Orkney, a 
Norse form signifying the " Isles of Ork." Old Gaelic ore 
signified a " pig, a whale;" hence " Whale-Isles " is the force 
of the word. Allied by root is the Lat. porcus, pig, whence 
radically we have the Eng. pork. The fourth island is the 
ever-famous Thule, but what the meaning of the name is or 
where the island was situated we do not here intend to discuss, 
for it is a very fruitless task. It is possibly part of Scandi- 
navia ; at least we cannot consider' Thule as belonging to 

In conclusion, I will now draw some inferences from these 
names given by Ptolemy for northern Scotland. Early Pict- 
land, we may take it, was Scotland north of the Firths of Forth 
and Clyde — the part of the country never subdued by Roman 
arms and called by Tacitus Caledonia. For this district north 
of the Firths up to the Orkneys, Ptolemy has gfiven us some 44 
names. Are the names Aryan by root and character ? Are 
they Celtic ? If Celtic, do they belong to the Brittonic or to 
the Gadelic branch of Celtic ? The first two questions are 
practically answered ; we have only to summarise the results at 
which we arrived in discussing the names separately. Of the 
44, three are translated names — Ripa Alta, Pinnata Castra, 
and Victoria ; these cannot count in our argument. The fol- 
lowing names we fooind (1) Celtic derivations for, and (2) 
noted them as existent either on the Continent in Celtic 
regions or in England and Wales on Brittonic ground, viz. : — 
Lemannonius, Itis, Tarvedum (Tarvedunum), Deva, Devana, 
Tava, Comavii, Decantse, Lugi, Smertse, Lindum, to which we 

1 Compare the goddess Domnu, whose name is in Inver Domnann 
(Rhys' Hib. Lect., p. 593). 

,'50 Ptolemy's giSography of Scotland. 

-may add Alauna. The root of Dumna and Dumnonii is 
common in Celtic lands, and the elements of Vaco-magi are 
easily paralleled in Gaul. Epidii is specially Brittonic, and 
good Celtic roots were found for Clota, Longos, Nabaros, 
CarnonacBS, Csereni, Vemicones, Orcades ; the Gaulish prefix 
ver appears in Vervedrum and Verubium; we suggested 
probable roots for Volsas, Ila, Varar, Loxa, Cselis, Cerones 
(Creones), Sketis, and also for Bannatia and Tamia. Even 
should our derivation of Caledonia be disputed, the root cal 
may be fallen back upon, and it is quite common in Celtic 
names ; but it is a root of several meanings. The Tvesis and 
Tina are doubtful as to form and origin (Spey and Eden ?) ; 
Bodotria, which is in a similar position, was referred to a 
Pictish comparative (vo-ter-) ; Orrea we referred to the root or, 
and Taexali was left underived. Skene suggest® for Orrea and 
Ila a Basque origin, a view that should commend itself to 
Professor Shys. 

We thus see that only three or four words cannot be satis- 
factorily accounted for ; and these, in two cases, are badly 
reoorded forms ; we are not sure that we are dealing with the 
genuine forms of the words. One-third of the names can 
easily be parallelel elsewhere on Celtic ground — Gaulish and 
Brittonic, but not, however, on Gadelic ground ; a fourth more 
show good Celtic rootsi and another fourth can be satisfactorily 
analysed into either Aryan or Celtic radicals. Hence we may 
justly conclude that the Picts or Caledonians spoke not only an 
Aryan, but also a Celtic language in the first century of our 
era. Two further facts point to the conclusion that the 
Pictish language and people were rather Brittonic than Gadelic. 
First, the p of Epidii is thoroughly non-Gaelic, but it is equally 
thoroughly Welsh; the root eqo or epo, as we saw, means 
" horse," and the former is the Gadelic and the latter the 
Brittonic form. Again, the names which are paralleled by 
■Gaulish and British similar forms clearly belong to Brittonic, 
or rather Gallo-British, ground, such as Devana, Tava, 
Alauna, Smertse, Itis ; these names cannot be got either in 
ancient or modem Ireland. We thus see that Ptolemy's 
geography of Pictland yields some proof that the Picts were, 
as to language, allied to the Cymric branch of the Celtic race. 
With later sources, such as Bede, Adamnan, the Annalists, and 
the Place-names, these proofs accumulate, so that now we may 
-claim, despite the cranky theories and objections of certain 
j)eople, that the Pictish question is settled. 


Journal devoted to the Literature, History, Antiquities, Folk 
Lore, Traditions, and the Social and Material Interests of 
the Celt at home and abroad. Conducted by Alexander 
Mackenzie (late Secretary of the Gaelic Society of Inverness), 
and the Rev. Alexander MacGregor, M.A., and Dr MacBain. 
13 vols. 10s 6d, each. 

The articles, etc., are those in wMch Highlanders take delight. — " John 
o' Groats Journal. 

No one can open its pages without finding something of pleasure, 
profit, or instruction. — " American Scotsman." 

The Magazine ought to find its way into the hands of every true 
Highlander. — " Greenock Highlander." 


Highlands of Scotland. As illustrated in the lives of John 
Farquharson and Alexander Davidson, the last of the Free- 
Foresters. By W. M'Combie Smith. 190 pages. 1904. 
Portraits and Illustrations. 3s 6d. 


John Farquharson — Lanavey — ^Lost on the Grampians — A Lesson in 
Deer-Stalking — ^Two Days with a Poacher of the Old School — Gamekeepers 
Outwitted — How the Deer were Carried Off — ^Initiating a Sportsman — 
Sportsmen and Keepers Sold — Seven Deer Shot in Sixty Seconds — Stalking 
Stalkers — ^Running the Blockade — A Dream and its Fulfilment — ^Fox Chasing 
— " Nell " — Alexander Davidson. 

Reads more like a, romance than a plain narrative of facts. — " Stirling 


Life. By David L. Cromb. Author of "The Highland 
Brigade : Its Battles and Its Heroes," _ etc. Nine Illustra- 
tions. 158 pages. 1903. 2s 6d. 


Alexander Macgregor, M.A. ; together with the History of 
the Feuds and Conflicts among the Claris in the Northern 
Parts of Scotland and in the Western Isles, from the year 
Mxxxi. unto MDCXix. First pubUshed in 1764 "from a MS. 
wrote in the reign of King James VI." 176 pages. 1907. 
3s 6d. 

MACKAY, Stirling. 


attributed to John Gait, Christopher North, etc. (Piano- 
forte) Music, Chrs. Helen Mackay. Is 6d net. 


(Coinneach Odhar Fiosaiche). By Alexander Mackenzie. 
2s 6d. 


with Genealogies of the Principal Families of the name, also 
the Families of Dunballoch and Phopachy. By Alexander 
Mackenzie, F.S.A. Scot. 25s ; Large Paper 42s. 


KENZIE. By Alexander Mackenzie.. Demy 8vo. 25s. 

ealogies of the Principal Families of the name. By Alex. 
Mackenzie, F.S.A. Scot. 25s ; Large Paper 42s. 


with genealogies of the Principal Families of the name, 
adding those of Lexington and New England. 25s ; Large 
Paper 42s. 


genealogies of the Principal Families of the name. 25s. 

BIBLE BEGINNINGS. By Rev. John Sinclair, 

M. A . ; with a short Preface by the Kev. Professor Sayce (of 
Oxford), D.D., LL.D. Grown 8vo. 5s. 


William Harvey. 3s 6d. 


Cameron, A.R.A. ; in Portfolio with descriptive letterpress. 
£2 2s. 

^: -*J- 

•?» "■ -I . i ^-'^ 



. V 

*v» „ 


mJ^. -■■■ 


r > ! • 

a •-i*-> ; 

'>*.w .