138 CRANIA BEITANNICA. [CHAP. V.
admit the co-existence in Gaiil of two languages, the one represented by modern Gaelic, the
other by the Cymric dialects *. This was the view of Niebuhr t ; which, ia a more detailed form,
has been so generally diffused by the work of M. Am. Thierry | . It is the opinion of this last writer
that the Gaelic language, which he terms the iirst or elder branch of the Celtic, was originally that
of the whole of GaiJ, and that, receding before the Cymric, or second branch, introduced by
Belgic tribes, it was gradually restricted to the central and southern regions; wMlst the Cymric
spread over the north of Gaul, including Armorica, and was thence communicated to the south
of Britain. This dualistic system, shaken by the main, results arrived at by Zeuss, now claims
fresh attention; and Pictet gives to it a qualified adhesion §.
Celtic philologists are generally agreed that the Gaelic and Cymric languages are derived
from one source, perhaps a common mother tongue of the Celtic now extinct; and the evidence
appeaa-s in favour of the Gaelic being the elder of the two ||, though even this has been disputed;
one writer maintaining that the Cymric was not only the original language of Britain, but also
that of Ireland, in which country the Irish was formed by separating from the Cymric This is
entirely at variance with the conclusions of the learned Edward Lhuyd, who found the roots of
the Celtic better preserved in the Irish than the Welsh; and who concludes from topographical
names in England, such as those of many rivers, as Usk, Esk, Ax, Ex, which are more readily
explained from the Irish iiisge (water), than from any Welsh word, that a dialect of the Gaelic
was the language of Britain before the introduction of the Cymric. This conclusion of Lhuyd's
coincides with that of Thierry formed on independent grounds, and in combination with it constitutes
a probable and consistent hypothesis **.
The relations of the Celtic to the other languages of the world, though indicated by Lhuyd,
were iirst systematically treated of by Prichard t t , who, on a comparison with the Sanscrit,
Greek, Latin and Teutonic, concluded that they belong to the great family called Indo-Germanic,
which, in addition to those already enumerated, comprises the Persian, Lithuanian, and Slavonic
* Pictet, "Inscriptions en Langue Gauloise," 1859, pp. 7,
52-54. These inscriptions present a new element for the decision
of this question, as they do not consist merely of proper
names, but contain verbs and pronouns arranged in the form
of brief sentences. See also M. Pictet's paper " On Irish and
Gaulish personal Names," Ulster Journal of Archseology,
vol. vii. p. 73.
t Lectures on Ancient Ethnology, Eng. ed. vol. ii. p. 305.
Eoman History, 1851, vol. ii. p. 522. Niehuhr's Lectures
were delivered before the appearance of Thierry's work, but
were not published until 1830.
X Hist, des Gaulois, 1828; and especially Introduction to
3rd (1844), 4th (1857), and 5th (1858) editions. Pictet
justly thinks that Thierry's system " errs by too absolute assertions."
§ Loc. cit. p. 54. In opposition to M. de Belloguet, who
says, (Ethuogéuie Gauloise, 1858, p. 282), "La dualité
Gauloise de M. Am. Thierry me paraît donc bien près de la
ruine au point de vue philologique." M. de Belloguet, in reference
to this question, divides Celtic scholars into Cymrists,
Irelandists and Bilinguists ; himself concluding that the ancient
Gaulish was scarcely separated into distinct dialects, but
was related at once to the Cymric and the Gaelic, more nearly
to the former by its vocabulary, and to the latter by the inflexions
which it still retained.
II J . Vf. Donaldson, "Cambridge Essays," 1856, pp. 33,
34, 39. Norris, loe. at. Sir G. Wilkinson, loc. at. p. 121. By
Dr. Donaldson, the existence of cases in the Gaelic is adduced
as evidence of its superior antiquity to the Welsh; whilst
Mr. Gamett curiously maintains that their absence in this
last " is a mark of antiquity exhibited by no other European
tongue in its original condition." Philological Essays, p. 204.
^ The views of Mr. Williams on this subject (loc. cit. and
prospectus of Cornish Dictionary) appear to follow those of
Gamett.
** Lhuyd's argument has been the subject of much criticism.
Gamett (Essays, p. 151) objects that though the Welsh wysff
does not mean water, it signifies a current or stream ; and concludes
that Lhuyd's hypothesis cannot be maintained. Mr.
Williams (loc. cit.) examines the other names adduced by
Lhuyd, and comes to a like conclusion. On the other hand. Dr.
Donaldson (Cambridge Essays, p. 39) and Sir G. Wilkinson
both adopt and even expand the argument. Dr. Prichard (vol. iii.
p. 151) on the whole concludes in favour of the hypothesis.
f j - Eastern Origin of the Celtic Nations, 1831. Compare
ante, p. 47. It was shown by Dr. Prichard that the Celtic
languages agree with those called Indo-German; I. in the
names of the numerals ; 2. in those of a certain proportion of
the commoner objects; and 3. in the personal terminations of
the verbs.
CHAP. V. ] mSTOBICAL ETHNOLOGY Or BEITAIN. 139
tongues. The conclusions of this writer were confirmed by the later researches of Pictet and
Bopp Prichard not only maintained the Indo-European alflnity of the Celtic dialects, but, as
the title of his book shows, the eastern origin of the people themselves; and this doctrine, m the
works of Max MMer and Bunsent, has been stiU further expanded. The Indo-European relationship
of the Celtic is not, however, universaUy admitted. Garnett was at first of opmion that
though Prichard had to a certain extent proved his point, yet "that the affinity between the
Celtic and the Indian family of languages is only p a r t i a l j D r . Latham admits the affimty of
the Celtic to the Indo-European languages, but denies the eastern origin of both language and
people; believing it more probable that the Sanscrit of India was derived from Eui-ope than that
nearly all the languages of Europe originated in Asia. He suggests that the Celtic, and mdeed
the Celts themselves, may have originated in Wales or Ireland; and demui-s to raising the value
of the Indo-European class of languages by the admission of the Celtic. He classifies the Celts
as " simple Europeans of the west of Europe; " or as " Occidental Japetidse," whom he separates
from the" Indo-Germanic JapetidiB," or those spealdng the other languages known as Indo-
European §. Notwithstanding these and other objections 1|, the Indo-European affinities of the
Celtic are now generally admitted. Dr. O'Donovan asserts that they are fully demonstrated by
Zeuss 1 , who without hesitation refers the Celtic to those " cognate languages which extend from
India through Asia and Europe to the extreme west **."
The important relation of the afanities of the Celtic languages to the problem, as to whence
the British islands were peopled, must be admitted; and by many they are regarded as the only
means for its solution. The sketch here given may serve to show the present state of the
question, the Unguistic bearings of which cannot be considered as definitively settled. If ever
decided, it must be, as Professor Pictet suggests, by the aid of fresh discoveries, and by a more
profound study of the remains of the Gaulish and of the ancient Irish and Welsh languages.
There is no evidence that the Celtic nations had an alphabet of their own in the Soman
period, or that their dialects could be properly classed with written languages. It is true that
in their public and private transactions the Gauls used Greek letters t t , an example of which are
the Usts of the numbers of the Helvetii, found by Cffisar in their camp J J. This use of HeUenic
* Pictet, " L'Affinité des Langues Celtiques, avec le Sanscrit,"
1837. Bopp, "Keltische Sprachen," 1839. See also
M. Pictet's late important work, " Les Origines Indo-Européennes
ou les Aryas Primitifs," 1859.
t " There was a time when the ancestors of the Celts, the
Germans, the Danes, the Greeks, the Italians, the Persians,
and Hindoos were living under the same roof separate from the
ancestors of the Semitic and Turanian races." Müller, " Oxford
Essays," 1856. See also his " Survey of Languages,"
and Bunsen, "Philosophy of Universal History," 1854.
t QuarterlyReview, Sept. 1836. Philol. Essays, pp.80,84.
After the appearance of Pictet's work, however, Garnett considered
the question as settled in the affirmative (p. 147).
§ EasternOrigin,2ndedit.pp.72,356,375,382. NaturalHistoryofVarietiesof
Man,1850,p.527. Norris's ed.of Prichard's
" Natural History of Man," 1855, p. 186. It is not perhaps
surprising that M. de Belloguet says of Dr. Latham, " il semble
* »admettre pour eux, (les Celtes) sans s'expliquer nettement,
je ne sais quel «liocAMomsme Âroj jéen" (loc.àt.f. 12). This
last expression is not employed by Dr. Latham ; who informs
the writer that " i t implies more of an hypothesis " than he holds.
II The Aryan or Indo-Germanic theory has lately been exammed
by Mr. J. Crawfurd, chiefly with respect to the resemblances
in the vocabularies. He conchides that the theory
which refers these languages with the people speakmg them to
a common origin is entirely groundless. (Memoir read before
the British Association, Oxford, 1860.)
^ Review of Zeuss's Gram. Celt. Ulster Journal of Archteology,
vol. vii. p. 11.
** Gram. Celt. pr. p. iii. M. de Belloguet considers the
Indo-European origin of the Celtic as a question definitively
solved (loc. cit. p. 8).
t t Csesar, B. G. Ub. vi. c. 14. " Grsecis utantur Uteris."
The Druids did not permit the use of letters in the education
of the youth of Gaul and Britain, and their discourses and
doctrines were committed to memory, not to writing.
i J B. G. lib. i. c. 29. "Tabulae Uteris Griecis confecta;."
The letter written by Cassar to Cicero (B. G. lib. v. c. 48) was,
for the sake of secresy, not merely in Greek characters, but in
the Greek language, as appears from Dion (lib. xl. § 9) and
PolyEenus, who gives the very words of Csesar. Stratag. lib. viii,
c. 23. § 6.