[2072] Keltic Researches, p. 16.
[2073] I find that, in the judgement of Prof. Haverfield (The Romanization of Roman Britain, p. 29), ‘the inscription ... may be best explained as the work of some Western Celt who reached Silchester before its British citizens abandoned it in despair.’
[2074] Chambers’s Encyclopaedia, 1901, vol. i, p. 279.
[2075] ‘In case,’ says Mr. Nicholson (Keltic Researches, p. 16, n. 2), ‘any one should quote against me Eppillus, the name of a son of Commius the Atrebat, as derived from epos for equos, let me say that in that case it ought to have only one p.’ No doubt it is remarkable that the p should be double (Rhys, Celtic Britain, 1904, p. 302); but Epillos, which is certainly the same word (A. Holder, Alt-celtischer Sprachschatz, i, 1445), occurs on Gallic coins of the Lemovices and on coins from Poitiers and the neighbourhood of Arles (E. Muret and M. A. Chabouillet, Cat. des monnaies gaul. de la Bibl. nat., 4578, 4579, 4580). See also Rev. celt., xxvi, 1905, p. 189.
[2076] Keltic Researches, pp. 17-8.
[2077] Tacitus, Ann., xiv, 31.
[2078] Ptolemy, Geogr., ii, 3, § 11.
[2079] Keltic Researches, pp. 25, 149. On page 26 (n. 1) Mr. Nicholson makes the curious suggestion that ‘the Britons, strictly speaking, were the Kymric branch who painted themselves, as distinguished from the Goidelic who tattooed’. Is he prepared to argue that the Belgae, who, on his theory, were Goidels, and with whom (p. 110) he apparently identifies ‘the original Brittones or Brittani’, were not included among ‘the Britons, strictly speaking’? Will he maintain, in the face of Caesar, from whom we learn that the Britons all ‘painted themselves (Omnes vero se Britanni vitro inficiunt)’, that ‘the great majority’ (p. 110) of the inhabitants of Britain did not paint? And, since Caesar undoubtedly included among the painted Britons the maritime tribes of the south-east, and also included them among the Belgae, does he not see the inconsistency into which he has fallen?
Bratuspantium, the name of a Belgic town mentioned by Caesar (B. G., ii, 13, § 2), would to most minds prove that the Belgae spoke a Gallo-Brythonic dialect, not only by the p which it contains, but also by the nt, a non-Goidelic combination. Mr. Nicholson, however (Keltic Researches, p. 16, with which cf. A. Holder, Alt-celtischer Sprachschatz, i, 515), of course explains the p as Indo-European.
[2080] It may be worth mentioning that Professor Rhys has affirmed (The Welsh People, p. 13) that the language of the British Goidels shows more traces of having been influenced by contact with the language of the non-Aryan aborigines than that of the Brythons. This fact, if it were a fact, would obviously be a further argument, if such were needed, against the view that the Goidels were the latest Celtic invaders of Britain. One expects, of course, to find that the professor changed this view, which was published in 1900; and accordingly we read in the address which he delivered in the same year to the British Association (Report, &c., p. 896) that ‘the syntax of insular Brythonic is no less non-Aryan than that of Goidelic’. Naturally in 1902 (The Welsh People, 3rd ed., p. 13) he repeated the former statement.