¹ Rodgers, R.H., L. Iuni Moderati Columellae Res rustica (Oxford, 2010), 406Google Scholar.

² The similarity of the two catalogues is noted by Boldrer, F., L. Iuni Moderati Columellae rei rusticae liber decimus (Pisa, 1996), 169Google Scholar, but without detailed discussion.

³ J.M. Gesner, Scriptores rei rusticae veteres latini, vol. 1 (Leipzig, 1735), 702.

⁴ The epithet Sarranae, of course, refers not to the geographical origin of the variety but to its colour (OLD s.v. Sarranus 2: ‘Dyed or coloured Tyrian purple’).

⁵ The attempt to establish the exact species implied by Columella is futile: see Boldrer (n. 2), 173, listing several alternative identifications; there is simply not enough detail in Columella's references, and it is not even certain that both species should belong to the genus Viola (Latin uiola could be used for flowers of quite other genera as well: see André, J., Les noms de plantes dans la Rome antique [Paris, 1985], 272Google Scholar; cf. id., Lexique des termes de botanique en latin [Paris, 1956], 330–1). Cf. in general Pliny's list of three varieties (HN 21.27): uiolis honos proximus, earumque plura genera, purpureae, luteae, albae, plantis omnes, ut olus, satae. For Columella's former variety, cf. Verg. Ecl. 2.47 pallentis uiolas, with W. Clausen, A Commentary on Virgil Eclogues (Oxford, 1994), 79: ‘pale yellow […], for the pallor of an Italian is not white or ashen but yellowish’; on pallere and related terms as implying a yellow hue rather than paleness, see further André, J., Etude sur les termes de couleur dans la langue latine (Paris, 1949), 139–47Google Scholar; cf. also R.G.M. Nisbet and N. Rudd, A Commentary on Horace, Odes, Book III (Oxford, 2004), 146.

⁶ Cf. Boldrer (n. 2), 173: ‘L'espressione purpurat auro indica secondo alcuni (che attribuiscono a purpuro semplice valore intensivo) un colore dorato.’ E. de Saint-Denis, Columelle: De l'agriculture, livre X (Paris, 1969) is somewhat ambiguous: in the commentary he says that the violet has ‘des fleurs de couleur or’ (at 56), whereas the translation speaks of ‘des fleurs d'un or rutilant’ (at 33‒4); cf. K. Ahrens, Columella: Über Lanswirtschaft (Berlin, 1972), 296, who translates: ‘in purpurnem Gold’.

⁷ See especially Edgeworth, R.J., mean, ‘Does “purpureus” “bright”?’, Glotta 57 (1979), 281‒91Google Scholar; cf. recently Librán-Moreno, M., ‘Blood-coloured swans: Hor. Carm. 4.1.10 and Homer's purple death’, CQ 67 (2017), 199‒209CrossRefGoogle Scholar, at 200. As concerns purpurare, the verb is fairly rare (see TLL 10.2.2714‒15), which itself speaks against its having developed a meaning different from that inherent in the stem. Perhaps one occurrence may require a brief comment, Apul. Met. 6.24 Horae rosis et ceteris floribus purpurabant omnia: this might suggest that purpurare means ‘to make bright, coloured’, but the mention of roses makes it clear that the red colour is meant in the first place (as the colour par excellence, as it were); cf. 10.29 purpureo nitore prata uestiret, where flowers of different colours could be implied, but likewise only roses are named (dirrupto spineo tegmine … promicarent rosae).

⁸ Note especially Hor. Epist. 2.1.207 lana Tarentino uiolas imitata ueneno and Ov. Met. 4.268–9 est in parte rubor uiolaeque simillimus ora | flos tegit, which show that the violet is naturally thought of as purple.

⁹ See e.g. Boldrer (n. 2), 173: ‘il verbo mantiene […] il proprio valore cromatico, offrendo con auro una combinazione di colori tipica della viola del pensiero o viola tricolor, in particolare della varietà hortensis con petali gialli macchiati al centro di purpureo’; but her translation ‘s'imporpora d'oro’ (at 65) sounds as ambiguous as purpurat auro. E.S. Foster and E.H. Heffner, Lucius Junius Moderatus Columella, On Agriculture X–XII; On Trees (Cambridge, Mass., 1955), 15 offer: ‘blooms with gold and purple blossoms crowned’. Similarly, W. Richter, Lucius Iunius Moderatus Columella; Zwölf Bücher über Landwirtschaft; Buch eines Unbekannten über Baumzüchtung, vol. 2 (Munich, 1982), 429: ‘purpurn und golden emporwachsen’. D.J. White, ‘Columella Res rustica 10: a study and commentary’ (Diss., Univ. of Florida, 2013), 74 translates (without explaining the grammar): ‘mixes purple with gold’.

¹⁰ Ursinus, F., Notae ad M. Catonem, M. Varronem, L. Columellam de re rustica (Rome, 1587), 182Google Scholar.

¹¹ In Burman, P., Titi Petronii Arbitri Satyricôn quae supersunt (Utrecht, 1709), 114Google Scholar. It is of course not entirely impossible to imagine auro being corrupted to abo, but the conjecture by no means imposes itself.

¹² Cf. Columella 12.18.5 aliter ea [sc. uasa] quae demersa sunt humi, aliter quae stant supra terram; and especially Arb. 4.1 uinearum autem fere genera in usu tria sunt: iugata, humi proiecta et deinde tertia … more arborum in se consistens.

¹³ For alte used of a flower, cf. Stat. Silv. 2.1.107 mollibus in pratis alte flos improbus exstat. I nevertheless have some reservations about the linguistic appropriateness of alte in this passage of Columella. In contrast to humi which expresses location in absolute terms (‘close to the ground’), and thus can be paired with a verb devoid of any connotation of place or direction such as pallere, alte expresses location only in relative terms and in reference to some other (implied) object, and I think for this reason it is predominantly used with verbs denoting motion or position (such as exstare in the Statius parallel), so that alte purpurare may not be good Latin.

¹⁴ Cf. e.g. 181 altera crispa uiret [sc. lactuca], fusco nitet altera crine, where the colour is first expressed with a verb (uiret), then with an adjective (fusco).

¹⁵ TLL 6.1.1347‒52 lists no instances of such usage. It has been suggested that frondes could refer to wreaths or garlands (OLD s.v. frons ¹ 2c), so that the expression would mean ‘the violet that (with its flowers) makes garlands purple’; however, this is merely a metonymical use of frons in its standard sense of ‘foliage’ for ‘a wreath made of (branches with) leaves’; and in any event, in a context speaking of a plant, frons can only refer to the plant's own leaves, not to some other leaves.

¹⁶ See e.g. Lundström, V., L. Iuni Moderati Columellae Rei rusticae liber decimus (Uppsala, 1942), 8Google Scholar; de Saint-Denis (n. 6), 33; Richter (n. 9), 428; Boldrer (n. 2), 64.

¹⁷ More or less the only context I know of where frondens is used of a flower (though, it may be noted, not when it is flowering) is Verg. G. 2.119 bacas semper frondentis acanthi, but here the point is that acanthus has evergreen leaves.

¹⁸ Bare frondens can normally be used of objects that are naturally thought of as having leaves, such as trees; thus arbor frondens would mean ‘tree in leaf’ (as opposed to ‘bare of leaves’, as trees can also be), which can further develop the pregnant sense ‘with lush foliage’. Since the notion of leaves is not an indispensable part of the concept of flower, bare frondens is not enough to convey a meaningful message about a flower's leaves. It may be instructive to compare Columella's other use of frondens in the poem, which occurs only ten lines after the present context (10.111): teneris frondens lactucula fibris, ‘lettuce green with tender leaves’; bare frondens would mean ‘when lettuce has full-grown leaves’ (as opposed to ‘when it has just been planted’, for instance), and it is necessary to add teneris … fibris to indicate what kind of leaves lettuce has.

¹⁹ See OLD s.v. scando 3b.

²⁰ See OLD s.v. surgo 8: the verb is predominantly used of plants sprouting from the ground (e.g. Prop. 1.20.37 surgebant lilia prato), but can also describe plants growing tall (e.g. Columella 6.23.2 fruticem surrecturum in altitudinem). In our context, its point will be secured by the opposition with humi.

²¹ Columella's treatise has other instances of a gloss interpolated either in the archetype (as here) or in one branch of the manuscript tradition. Of the former type, perhaps the clearest case is 11.1.14 ueneriis amoribus for ueneriis rebus (Schneider), where amoribus is a gloss on the whole idiom but only replaces its second part. Other examples include (but are not limited to): 7.3.24 sidus aestatis per emersum Caniculae, where sidus aestatis (del. Hine) is a gloss on Caniculae; 8.5.3 pulicibus atque aliis similibus animalibus (animalibus om. AS), where similibus (del. Rodgers) is an exegetical gloss on aliis animalibus, which has replaced animalibus in AS but has been added to it in R; 9.13.2 samaras ulmi semina sua promunt, where semina sua (del. Iucundus) is a gloss on samaras; 11.2.52 proscissum ueruactum, where proscissum (del. Rodgers) is a gloss on ueruactum.

²² We may compare the reverse corruption at Ciris 184: oestro AR, estro H, ostro ρ.

²³ Alternatively, the gloss may have been tabāo, which was then changed to tabo for the sake of metre.

²⁴ This remark appears close to the end of the book immediately preceding Columella's poem, which makes it the more likely that the term could still be in the reader's mind when he came to 10.101.

²⁵ Moduin's Eclogues survive in two manuscripts, both from the ninth or tenth century (British Library Add. 11034 and Darmstadt, Universitäts- und Landesbibliothek 3301). The first editor printed tabano (he published three editions, of which only the last was based on both manuscripts): Dümmler, E., ‘Gedichte des Naso’, Zeitschrift für deutsches Alterthum 18 (1875), 58–70Google Scholar, at 63; id., ‘Nasonis (Muaduuini) ecloga’, in id., Poetae Latini aevi Carolini, vol. 1 (Berlin, 1881), 382–92, at 388; id., ‘Nasos (Modoins) Gedichte an Karl den Grossen’, Neues Archiv der Gesellschaft für ältere deutsche Geschichtskunde 11 (1886), 75–91, at 86. But see Traube, L., ‘Bombo. tabo’, Archiv für lateinische Lexikographie und Grammatik 16 (1889), 167–8Google Scholar, at 168, defending the transmitted form. In any event, it is of little importance for the present argument whether tabone is the original reading or merely the reading of the archetype, as in either case it attests to the existence of the form in medieval Latin.