¹ Hesiod, , W. and D. 766 ff.Google Scholar gives a number of clear examples, e.g., looking over work and dealing out supplies (1. 766); shearing sheep and reaping fruits (ll. 774–5); being bom (ll. 782, 785); gelding kids and sheep (l. 785); bringing home a bride (1. 800).

² e.g. beginning to sow (Hesiod, , W. and D. 780–1Google Scholar); beginning to build narrow ships (l. 809).

³ e.g. in Hesiod, , W. and D. 780–1Google Scholar, the thirteenth day is mentioned as unlucky for beginning to sow, yet the best day for setting plants. Clearly setting plants is an operation that may last a number of days, like sowing; hence it is the beginning of the operation to which Hesiod refers.

⁴ Cf. Blümner, H., Technologie u. Terminologie der Gewerbe u. Künste bei Griechen u. Römern, i ² (1912), 144, n. 3.Google ScholarErnout-Meillet, , Dict. Etym.² (1939)Google Scholar, s.v. ‘ordiri’, regards the technical meaning as the original one, the general sense being an extension, due to a confusion of ordior and orior. A less likely view makes the general meaning the primary one. It is sufficient for my purpose to show that the attaching of the warp-threads to the beam was the generally accepted exordium of weaving.

¹ See Blümner, , 143–4Google Scholar, for references. The Attic word was ἄττεσθαι.

² For the translation of tela as ‘loom’ see below, p. 96, n. 3, and p. 100. For this sense of intendere cf. Virg. Aen. iv. 506Google Scholar: ‘intenditque locum sertis’—Dido hangs the place with garlands. In the Ovidian context there is, of course, the further idea of ‘tautness’ in the hanging of the warp.

³ So MSS. I have not accepted Merkel's emendation Pallade telae.

⁴ With this text F. J. Miller translates (Loeb ed.): ‘she hangs a Thracian web on her loom’; I do not understand this. For further examples of this operation of exordiri see Tibull. i. 6. 79Google Scholar; Epithal. Laurent. (Reise, Anth. Lat. i. 2. 211, no. 742) 44 (discussed below).Google Scholar

⁵ Op. cit. 143, n. 3. Heyne, too, commenting ad loc. considers this possibility: but he refuses to commit himself, writing: ‘addere licia telis; est telam ordiri: ut adeo tela sit sive stamen et trama, sive ipsa machina textoria cui alligatur’.

⁶ Aug. Hug, P.-W., s.v. ‘tela’, col. 177.

⁷ Ibid., col. 174.

⁸ The same definition is repeated in Forcellini-Corradini-Perin.

⁹ The OED defines ‘thrum’ as: ‘the ends of the warp-threads left unwoven and remaining attached to the loom when the web is cut off’. This is not quite the sense in which the word is being used by Forcellini-Bailey.

¹⁰ Sidgwick, however (ed. 1896), adopts it in his note ad loc. For its possible origin, see below, p. 97, n. 1.

¹¹ A heald consists of a wooden frame containing a number of vertical wires, known as leashes or heddles, each with a ‘mail’ in the middle, through which a warp-thread can be passed. On a hand-loom the raising and lowering of the healds is effected by means of a foot-treadle; but on a power-loom, particularly in the weaving of complicated patterns, a more involved system of levers is frequently employed, controlled through a series of perforated pattern-cards fitting over needles.

¹ See below. A multiplication of healds is only feasible with the introduction of the horizontal loom: and it seems unlikely that the Greeks and Romans of the classical period possessed this (cf. Blümner, , 137 ff.).Google Scholar

² It is also to be found in Georges, Lat.-deutsch. Handwörterb.⁸ One need not pause over the view expressed in the seventh edition of this work and in Walde, Etym. Wörterb. (abandoned in Walde-Hofmann, ed. 3), that licium means ‘weft’. Neither quotes any example which is clarified by giving the word such a meaning.

³ The same version is found in Conington-Nettleship⁴. Originally Conington had translated ‘to add the leashes of the woof to the warp’, which is meaningless; he may have been led astray by Papillon's comment: ‘“leashes” or “thrums” for attaching the threads of the warp (tela) to the woof (sic!); licia telae addere = “to weave”’.

⁴ On the construction of the ancient loom see Blümner, , 135 ff.Google Scholar; Johl, C. H., Die Webstühle der Griechen u. Römer (Kiel Diss.), Leipzig, 1917Google Scholar, with bibliography; Aug. Hug, P.-W., s.v. ‘tela’; Wilson, Lilian M., Ancient Textiles from Egypt in the University of Michigan Collection (1933), 4–8.Google Scholar

⁵ With a patterned cloth the process is, of course, more complicated.

⁶ For diagrams see Blümner, , 156Google Scholar; Johl, 24; Wilson, 6 (erratum slip opposite).

⁷ Ovid, Met. vi. 55Google Scholar: ‘stamen secernit arundo’.

⁸ This rod is the κανών of Homer, Iliad, xxiii. 761.Google Scholar

⁹ e.g. Rhoades, J.: ‘and fix the leashes to the warp’Google Scholar; Goelzer, H., ‘pour mettre de nouvelles lisses à la chaîne’ (Budé trans.)Google Scholar; Way, A. S.: ‘and for tying the loops to the warp’Google Scholar; Fairclough, H. R., ‘and for adding the leashes to the warp’ (Loeb trans.)Google Scholar; Ernout-Meillet, , Dict. Etym.² s.v. ‘licium’Google Scholar, defines the word as a leash, and quotes this passage in support; there are probably other examples.

¹ The brackets indicate works which quote each passage in the sense of ‘warp’.

² For Ovid, Fasti, iii. 819Google Scholar, which also contains the phrase stantes telae, see below, n. 6.

³ See, for l. 54, above, p. 94. In l. 55: tela iugo vincta (al.: iuncta) est apparently means ‘the loom was fastened together by means of the beam (i.e. the warp-beam)’; cf. Lafaye, G. (Budé trans.) ‘le métier est joint à l'aide de la traverse’.Google Scholar Miller (Loeb trans.), however, translates ‘the web is bound upon the beam’. Presumably by ‘web’ he means ‘warp’ (for the actual weaving had not yet begun); but in any case the use of tela in different senses in ll. 54 and 55 involves some difficulty.

⁴ Cf. Conington-Nettleship, ad loc.

⁵ Forcellini-Corradini-Perin quotes the same three examples as Forcellini-Bailey.

⁶ Based on Georg, i. 294Google Scholar, where, however, there s no adjective qualifying telas to settle the meaning one way or the other; this is true too of the echo in Ovid, Fasti, iii. 819Google Scholar: ‘stantis radio percurrere telas’, where the editors generally give to tela the meaning ‘warp’ rather than ‘loom’. So both Bailey and Frazer understand it. In Ausonius, Mosella, 397Google Scholar: ‘Pierides tenui … aptas subtemine telas percurrent’, aptas could apply equally well to an ‘assembled loom’ or a ‘warp mounted (on the beam)’; subtemine will go with percurrent, not with aptas.

⁷ Page argues that argutus is more applicable to the shuttle than to the reed (cf. Arisi. Frogs, 1316: κερκὶς ἀοιδός); and he points to the similar passage, Ovid, Fasti, iii. 819Google Scholar (see previous note), where the radius is undoubtedly the shuttle.

⁸ For the same operation cf. Ovid, Fasti, iii. 820Google Scholar: ‘rarum pectine denset opus.’ See Blümner, , 159.Google Scholar

⁹ For tela in the well-attested sense of ‘cloth’, see, e.g., Virgil, , Aen. iv. 264Google Scholar; xi. 75; Georg, iii. 562Google Scholar; CGL, iii. 270Google Scholar: ὅøασμα: pannus, tela.

¹⁰ Another case of tela signifying ‘warp’ is in Epithal. Laurent. 45–6Google Scholar (discussed below, p. 98, n. 1); see too CGL, v. 580.Google Scholar 54 (= Isid. Etym. xix. 29. 1): ‘tela: pro longitudine staminum’.

¹¹ This etymology, which is quite certain in view of the French lisse and the Italian liccio, is not listed in the OED, which derives this, like the other uses of ‘leash’, from ‘OF “lesse”, perh. from fern, of Latin laxus’. The alternative spelling ‘leish’ for the technical sense also points to a separate origin.

¹ He probably found it in Donatus, Aelius, ad Ter. Andr. 911Google Scholar: ‘licia enim dicta sunt quasi ligia’. It is not impossible that these two passages are behind the curious idea of licia contained in Forcellini-Bailey (see above, p. 94).

² For the meaning of telas in this passage, see below, p. 98, n. 1.

³ If licia is leashes, telis may be ‘loom’ (as we shall see); but equally well it can be ‘warp’. The use of the plural telis has no special significance.

⁴ On the primitive loom, it is worth noting, the fastening of the leashes to the warp was subsequent to its being mounted on the beam; see Crowfoot, and Griffiths, , JEA, xxv, 1939, 46.Google Scholar

¹ Blümner sees a further parallel in the passage Epithal. Laurent. 45–6:Google Scholar

compositas tenui suspendis stamine telas;

quas cum multiplica frenarint licia gressu …;

but here closer examination suggests another meaning. Blümner evidently interprets l. 45 as ‘you suspend the slender warp from the assembled loom’. But against this (1) quas in l. 46 must mean the same as telas in l. 45: and it is the warp rather than the loom that is ‘curbed’ by the leashes; (2) the construction suspendo lelas stamine has no parallel, the usual form being suspendo stamen telis. Attractive therefore though the parallel to Tibull. i. 6. 78 would be, I think one must translate: ‘you suspend the warp composed of slender thread’. For this sense of compositus cf. Sil. Ital. xv. 51: ‘mensam gramine compositam’.

² Cesareo's version (ed. 1938): ‘e unisce tenaci licci nello stretto ordito’ apparently refers the passage to the actual process of weaving (if licci is ‘weft’ and ordito ‘warp’).

³ On this technique see Blümner, , 169 ff., 219, n. 4.Google Scholar

⁴ Francken, ad loc., comments: ‘miscere licia: coniungere subtemini diversi colons’, and shortly afterwards: ‘licia … Aegyptia versicolora fuerunt’. His meaning is not clear: what one ‘joins to a weft of a different colour’ are presumably the warp-threads (though the reverse expression would be more accurate); but if it is the Egyptian warp-threads that are versicolora, the only literary authority for such a view is Pliny, , viii. 196Google Scholar (discussed below, pp. 101 ff.), in which, as we shall see, there is considerable difficulty in taking plurimis liciis to mean ‘with multi-coloured warp-threads’. Perhaps Francken is using licia in the sense of ‘threads’ in general.

⁵ Below, pp. 100, 101.

⁶ This is preferable to joining Phariis with telis.

⁷ Bourgery and Ponchont appear to adopt this general meaning in their Budé translation: ‘dans la manière artistique qu'ont les Égyptiens d'ourdir leur tissus.’ Their use of ourdir (‘to warp’, from Latin ordirt) is significant for my argument.

⁸ That multi-coloured, patterned textiles were common in Egypt at a not very much later date than this is clear from the collections of Graeco-Roman and Coptic textile fragments in the museums (cf. Wilson, op. cit.; A. F. Kendrick, Catalogue of Textiles from Burying Grounds in Egypt (Victoria and Albert Museum)). But it is very rare for the warp to play any part in the pattern of such materials as we possess.

¹ For licium meaning ‘thread’ (not further specified) see below, p. 100.

² Bergman includes the Virgil reference in his Index Imitationum; and there are two other certain Virgilian reminiscences at ll. 1108 (cf. Aen. xi. 199 and xii. 214Google Scholar) and 1113 (cf. Georg. iii. 171).Google Scholar

³ On this subject see Daremberg-Saglio, iii. 515, s.v. ‘Infula’; v. 759, s.v. ‘Vestales’.

⁴ Servius, ad Aen. x. 538Google Scholar: ‘Infula, fascia in modum diadematis a qua vittae ab utraque parte dependent, quae plerumque lata est, plerumque tortilis, de albo et cocco’.

⁵ Festus, 393 a.

⁶ Examples of this Virgilianism are to be found quoted on almost any page of Henry's Aeneidea.

⁷ For the equation licium = μίτος see below, p. 101, n. 4.

¹ One should distinguish, I think, between the fortuitous ambiguity of Tibullus and the deliberate vagueness of Lucan.

² I omit all consideration of those meanings of licia without obvious relevance to weaving or cloth as such.

³ For tela, ‘cloth’ see above, p. 96, n. 9.

⁴ This does not mean that the loom is made of iron, but that iron tools are used in its preparation—which is not true of the warp.

⁵ Cf. Blümner, , 146Google Scholar; Johl, 19 ff. Blümner, , 158–9Google Scholar, believes Seneca to be contrasting two forms of loom, the earlier with weights, on which one wove from above downwards, and the later type in which both ends of the warp were fastened to beams, and the weaving was from below upwards. Summers, Select Letters of Seneca, rightly points out that this is not the logical sense of the passage, Seneca being concerned, not with later improvements to the loom, but with whether the original invention was to be attributed (as Posidonius claims) to the philosophers.

⁶ Here strictoria de tela is apparently a shirt ‘straight from the loom’, i.e. unworn.

⁷ References in CGL, vii. 335, s.v. ‘tela’.Google Scholar

⁸ Op. cit. 141, n. 8. In two passages of Ausonius, Epig. 38Google Scholar (55 Peip.), 1, and Epist. 23 (28 Peip.), 14Google Scholar, Forcellini-Bailey (followed by Forcellini-Corradini-Perin) claims that ‘licia dicuntur ipsa stamina, quibus tela texitur’. But in both cases the meaning of licia seems to be ‘cloth’; cf. Blümner, , 142, n. 1.Google Scholar

⁹ See below, Addendum, , pp. 101 ff.Google Scholar

¹⁰ Cf. too Lucan, x. 126Google Scholar (above, pp. 98–9); Amm. Marc. xiv. 6. 9Google Scholar: ‘tunicae … varietale liciorum effigiatae in species animalium multiformes’ (where the figures of animals can only have been obtained by a variety either in all the threads, both warp and weft, or—if tapestry weaving is referred to, in which the weft is beaten up to cover the warp almost entirely—in the weft. It is of course possible that the reference is to embroidery, not weaving).

¹¹ Above, pp. 98–9.

¹² Line 681 (Marx) = 721 (Terzaghi) = 638 (Warmington).

¹ e.g. Buecheler, , Rhein. Mus. xliii. 1888, 291 ff.Google Scholar punctuates ‘laterem in telam, licium’, and takes laterem as the same as ἀγνθα, ‘loomweight’. F. Marx, ad loc., compares Tibull. ii. 1. 66: ‘et a pulso tela sonat latere’, from which he deduces that later is the already woven part of the cloth, which is beaten tight with the pecten (see above, p. 96).

² Marx has a note which leaves it quite uncertain what he takes licium to be; Warmington, punctuating ‘in laterem in telam licium’, translates ‘a thrum for the wool-ball and for the warp’!

³ As quoted by Blümner, , 141Google Scholar, n. 8, this passage is untranslatable: the text is very uncertain.

⁴ The examples are listed in CGL, vi. 644.Google Scholar

⁵ e.g. Homer, , Iliad, xxiii. 762Google Scholar (cf. Blümner, , 148–9)Google Scholar; Schol. ad loc. κάλαμος, περὶ ὃν εἱλεται ὁ μίτος ὁ ἱστουργικός; Anth. Pal. vi. 174. 6Google Scholar; Hesych. s.v. κανών τὸ ξύλον περὶ ὅ ὁ μίτος. Blümner, , 149, n. 1Google Scholar, suggests that μίτος is the set-up warp, στήμων the thread employed to build it.

⁶ There is an interesting, if not very decisive, discussion of polymita in Salmasius's edition (1620) of the Historia Augusta, pp. 310 A-D, 356 B-F, 396 C-D, 511 D-512 F.Google Scholar

⁷ μίτος is not found in the sense of ‘leash’; but as licium acquired that meaning (above, pp. 96–7) it seems likely that μίτος did so too. See too below, p. 103, n. 1, for the modern Cretan use of μίτος as ‘heald’, an extended meaning which logically demands the intermediate sense of ‘leash’. Further, as Johl, , 29–30Google Scholar, describes, the leashes on a form of modern Greek loom are known as μιτάρια.

¹ Crowfoot, and Griffiths, , JEA, xxv, 1939, 47.Google Scholar

² e.g. CGL, v. 452Google Scholar: ‘κάνναβις, μίτος, νμα, filum: inde dicta vestis polimita quae panno tenuiorum filorum constat, qui multitudinem filorum continet bene textorum’.

³ e.g. CGL, v. 234, 31Google Scholar: ‘polemos, bellum, pugna, graece. polemitum ergo quasi pugnas textas’.

⁴ The etymology of multicius is uncertain. Leumann, , Glotta, ix, 1918, 152 ff.Google Scholar, derives it from *multus, a past part. pass, of mulceo; Salmasius, , ad Valer. Aug. apud Vop. Aurel. 12 (p. 356 B)Google Scholar, takes it from multus-icio, Gk. πολνσπαθητόν; so too Georges, , Lat.-deutch. Wörterb ⁸Google Scholar. However, the ancient grammarians held it to be a contraction from multilicius: cf. CGL, v. 524. 7Google Scholar; 573. 13: ‘multitia vestis quae multa licia habet’; Val. Aug. apud Vop. Aurel. 12Google Scholar, where MSS. give both multicias and multilicias. Both Walde-Hofmann, , Etym. Wörterb.³s.v. ‘licium’Google Scholar and Ernout-Meillet, , Dict. Etym.²Google Scholar, accept this identification as probable, and Ernout-Meillet suggests that the corruption arose from the analogy of such adjective forms as emptus/emplicius, novus/novicius.

⁵ Cf. CGL, v. 653. 5Google Scholar: ‘multitia genus vestis pluribus coloribus confecta’; v. 576. 23: ‘polymita: multicoloria, varia’ (cf. v. 524. 32); v. 382. 7: ‘polimita hring faag’ (Ang.-Sax. meaning, apparently, ‘with circular bands of different colours’); v. 524. 34; 576. 29: ‘polymatus est textus multorum colorum’ (cf. Isid. Etymol. xix. 22. 21Google Scholar: ‘polymita multi coloris: polymitus enim textus multorum colorum est’); Du Cange, , Gloss, s.v. ‘polymitus’: ‘Polymita vestis, multis variisque coloris filis et liciis contexta et variegata, apud Jul. Pollucem’Google Scholar. Gloss. MSS. in lib. Job et Phavorinus: ποικιλτική, πολνμιταρική. Gloss. Saxon. Aeljrici: ‘polimita vel oculata: hringfegh’.’ Further search in medieval glossaries would no doubt produce further examples.

⁶ Op. cit. 142: ‘buntgewirkte Stoffe, bei denen verschiedenfarbige Kettenfäden zur Verwendung kommen’.

⁷ A technical expression is hardly likely to have perpetuated the kind of error that, on our first interpretation of the passage, Lucan made in x. 126 (see above, pp. 98–9). However, if miscendi licia in that passage is to be taken as ‘mixing the warp-threads’, Lucan's technical mistake probably arose because he knew that polymita was the name of the multi-coloured stuff he was describing, and that this name should mean ‘with many warp-threads’, and hence assumed that the variety of colour was obtained in the warp. If this is so, it offers further striking confirmation of the thesis that licia commonly meant ‘warp-threads’.

⁸ According to Johl, , 52Google Scholar, and Blümner, , 168Google Scholar, this describes a form of tapestry weaving. That the Egyptian cloths contained some new feature is, however, suggested by Martial, , xiv. 150Google Scholar (entitled cubicularia polymita):

Haec tibi Memphitis tellus dat munera: victa est pectine Niliaco iam Babylonos acus.

Johl argues that the change was merely one of pattern and fineness of weave.

¹ Crowfoot, and Griffiths, , op. cit. 47.Google Scholar

² For this, the earlier Latin word for a heald, see Epinal Gloss. 602Google Scholar: ‘liciatorium, hebild (a.d. 700)’; cf. Corpus Glossary (ed. Hesseis, ), L. 178Google Scholar: ‘liciatorium, hebelgerd’. The extended meaning of licium is seen in Corpus Glossary, L. 249–50: ‘licium, hebeld; licia, hebeldðred (c. a.d. 725)’Google Scholar; Suppl. Aelfric's Voc. in Wright-Wülcker, , 187, n. 13: ‘licium, hefeld’Google Scholar; cf. Napier, , Old English Glosses, 3545, 3550Google Scholar; 2. 222; 4. 65; 7. 256.

³ OED, s.v. ‘heald’.

⁴ Marquardt-Mau, , Das Privatleben der Römer (1886), 524.Google Scholar

⁵ Op. cit. ed. 1, 142.

⁶ Op. cit. ed. 2, 141, n. 7: ‘Dass μίτος das Geschirr am horizontalen Webstuhl bedeute … beruhte auf unrichtiger Interpretation’.

⁷ e.g. in Lewis and Short.

⁸ OED: from OE. twili, cogn. with OHG. zwilîh, after Latin bilix from licium. Cf. CGL, ii. 10 (corr. a): ‘bilix vestis duplici licio vel lorica, tuili (Ang.-Sax.)’Google Scholar. (For a bilix lorica cf. Virgil, , Aen. xii. 375.)Google Scholar

⁹ From the low Latin dimitum, Gk. δίμιτος, perhaps via Italian. Originally it was the same as bilix; cf. CGL, ii. 29. 56Google Scholar: ‘bilex, δίμιτος’; ii. 277. 59; iii. 276. 35 (given under the heading de legibus!): ‘δίμιτος, bilex’; iii. 322. 48: ‘δίμιτον, bilicae’; iii. 514. 42: ‘dimiton, hiligce’; but today it signifies a stout cotton cloth with raised stripes and fancy figures used until recently for bedroom hangings.

¹⁰ From ‘drilling’, from German drillich, from Latin trilicem, nom. trilix. Trilix (like bilix) is used by Virgil of a leather breastplate (cf. Aen. iii. 467Google Scholar; v. 259; vii. 639; Sil. Ital. ii. 401Google Scholar), which suggests a special sense; cf. too Isidor. Elymol. xix. 22. 23Google Scholar: ‘trilicis a tribus liciis, quia est et simplex et bilex’. Does this refer to some kind of alternation between a plain and a twill weave? or to a simple weave with a double warp? We simply do not know. The modern meaning of ‘drill’ is ‘a coarse twilled linen or cotton fabric’.

¹¹ Obsolete: in the Middle Ages it was a rich silk fabric; cf. Salmasius, , edit: Hist. Aug. 512 B. Med. Gk.: ἑξάμιτοςGoogle Scholar. Middleton, , Enc. Brit.¹⁰ xxiii. 201–2Google Scholar (quoted in OED, s.v. ‘samite’) says: ‘the weft threads were only caught and looped at every sixth thread of the warp, lying loosely on the intermediate part’. Du Cange, , Gloss, s.v. ‘dimitum’Google Scholar, gives a 12thcent. Latin quotation describing a cloth factory where you could see (as well as exhimita) amita dimitaque et trimita—which the OED translates as ‘fabrics woven with one (sic!), two, or three threads respectively’. Per obscurum obscurius! 12 See above, n. 8.

¹ From examples taken from the Navaho Indians, Ceylon, and elsewhere, Mmes Crowfoot and Griffiths show, however (op. cit. 43, 46), that quite complicated materials are capable of fabrication on extremely primitive looms.

² See above, p. 102, n. 8.

³ Pliny's plurimis liciis texere is, of course, simply a literal rendering of polymita; whether these words conveyed any clear technical process to Pliny's mind, and how he meant licia to be understood, remains uncertain. There is, however, no evidence that licium, any more than μίτος, could mean ‘heald’ as early as the first century a.d.

⁴ I am grateful to Professor J. F. Mountford for a number of valuable hints and suggestions, which have been incorporated at various points in the argument, and to Mr. R. J. Getty for the references to Forcellini-Corradini-Perin.