Published online by Cambridge University Press: 22 April 2026
Scholars follow error as willingly as truth, and accept opinion as readily as fact. A sentence falsely ascribed to Columella by Holder in his Alceltischer Sprachschatz (2.1245, 51-3) is to be found in no manuscript or printed edition of Columella. Whence Holder got it, remains a mystery, but at least twenty scholars have repeated it (from Holder) and ascribed it to Columella without stopping to verify the text (cf. PID 2.204). An object discovered at Gièvres (in the Loir-et-Cher) in the middle of the last century was described as an amulet and the inscription which it bears misread when first published. Both the spurious description and the erroneous transcription were accepted by Hirschfeld (CIL 13.1324) and repeated by Holder (op.cit. 2.1011, 54-1012.2), who writes
pixtionovimxmorucin
n which x before t, if the reading were justified, would have to be transcribed χ, as in Pixtacus, Pixtaucus (with a : au, see my paper in the Havers Festschrift, Die Sprache 1, 1949), Pixticenus (i.e. -genus), Pixtil(l)us beside Pictaui, Picones, Pictil(l)us, and Picti. As for x before (and after) m, that is quite vapid.
1 147 in my forthcoming Dialects of Ancient Gaul (DAG).
2 DAG, Note xxxi (g).
3 Bulletin de la Société archéologique, scientifique et littéraire du Vendômais 11.102, plate 2, fig. 2 (1872).
4 Bulletin archéologique (du Comité des travaux historiques et archéologiques) 1914.214.
5 Ibid. 221, Nos. 1 and 8.
6 DAG, Note xxxi (a, e, g), all found at Aûtun.
7 DAG 200 (Eschweilerhof, in Gallia Belgica), 228 (Blickweiler, in Germania Superior).
8 DAG 83, 214, 224, 244, 250.
9 DAG 164.
10 On these see K. Bücher, Arbeit und Rhythmus 42-3 (Leipzig, 1896). The refrain in Catullus 64 (e.g. 327; cf. Verg. Eel. 4.46), currite ducentes subtegmina (‘threads’), currite, fusi (‘spindles’) is in point; or the well-known adespoton (Bergk, Poetae lyrici graeci III 672, fr. 43)
in which
is vocative, like fusi in Catullus. In both of these the address is to the implement, the spindles or the mill. That fate is involved in spinning is an ancient commonplace, of which I need to remind no one.
11 Walde-Pokorny 1.223
- ‘drehen, biegen’; cf. 227 on ‘extensions’ (or
(223-3, 240-1).
12 Morris-Jones, Welsh Grammar 236, 239 (1913). I need hardly say that I am thoroughly familiar with Loth’s discussion of uimpi ‘gwymp’ in Comtes rendus (Académie des Inscriptions et Belles-Lettres) 168-86 (1916).
13 DAG 178.
14 DAG 158.
15 DAG 164. Cf. Gk.
, Arm. im, Alb. im? The present essay in interpretation is offered to Roland G. Kent as a conspicuous and successful exponent of the method here used, a combination of the philological with the linguistic.
16 Walde-Pokorny 2.621.