Tablet 596

description 182 x 66 mm. Bowman and Thomas (1996), no.1, Plate VII. Plate 6.
Archaeological data. Location: Site II/NW. Period: 3?.
A complete diptych, discovered close to the south gate in the Period 3 context. There are notches in the right- and left-hand edges which establish the relative positions of the two halves of the diptych and the sequence of the text. The tablet contains, on one side, two columns of a text written along the grain of the wood. The left-hand half of the diptych has remains of writing on the other side; this is much less well-preserved and we therefore suppose that it was the outer face or back of the folded diptych. The back of the right-hand half is blank. The text on the back is certainly part of a similar document, probably, although not necessarily, the same as that on the front. This is perhaps a little surprising in view of the fact that letter-writers normally write the address on the back of the right-hand half of a folded diptych, but there is no way of being sure whether what is on the outside comes before or after what is on the inside. We did not attempt a transcription in the; the scan has enabled us to make out some of the writing but we have not been able to elucidate it satisfactorily (see notes). The format on the front is precisely comparable with that of 182. It is notable that the beginning of the present text (like 182) contains no sign of a date heading. It may be that it was simply undated or that the text began on another tablet which has not survived.
At first sight, the text is most obviously to be taken as an account, but it is a curious miscellany which includes items of clothing, some equestrian equipment, domestic utensils and textiles. There is no obvious single category into which all the items might fall and this perhaps suggests a private memorandum, or perhaps a simple inventory or "asset register" with valuations of items purchased for the praetorium and its inhabitants by one of the people (slaves?) responsible for the domestic administration. This would perhaps explain better the presence in all entries of unit prices and totals. A comparable list, including two of the same items, may be found in P.Freib.2 = CPL 314 = ChLA XI 485 (second century AD), an inventory of silver jewellery and other personal items. We have considered the possibility that there might be more references to equestrian equipment than is immediately obvious: i.e. that clausurae, infiblatoria, capitularia, lumbaria and uela could be interpreted in this context; but even if this were admitted, we would still be left with saga corticia and trullae defying such a categorisation. It might be better to think of some of the items in the context of luxury goods or jewellery belonging to women, but, again, this will not account for all the items listed.
The text has a number of features of particular importance. First, there is a variety of unusual items, some of which are designated by terms of lexicographical interest. Second, the presence of unit prices as well as totals for the multiple items in the account, yields information which is of particular importance for economic historians; there are only a few other examples of this in the Vindolanda texts, one of which is illegible (185.27-28, 192.4-5, on which see below, and perhaps 582), cf. Introduction, pp.00. Third, the account has several occurrences of fractional signs designating divisions of the denarius. Examples of these may be found in several other accounts including 182, 184, 193 and 588 (see also Tab.Vindol.II, pp.54-5); when the texts in Tab.Vindol.II were published, however, we did not have enough evidence to be confident of their meaning. Some of these symbols may be the same as those given in Volusius Maecianus, 47-62 (= Metrologicorum Scriptorum Reliquiae, ed. F. Hultsch, II, pp. 67-8), though we cannot be sure how far the printed symbols in the Teubner edition accurately reflect what is contained in the MSS. Even if some of the symbols are the same, however, others, as far as we can ascertain, seem to have occurred only in the texts from Vindolanda. They cannot be expected to occur in comparable Latin accounts on papyri from Egypt where an idiosyncratic system of using the local obol unit as the fraction of the denarius was used.
The fact that this account offers unit prices and multiplied totals enables us to check the arithmetic and to identify with some confidence the symbols used to represent a quarter-denarius (= 4 asses), an eighth of a denarius (= 2 asses) and a sixteenth of a denarius (= 1 as) respectively. In this system, the quarter is represented by a horizontal dash with a vertical beneath it curling to the left [drawing]. The eighth is a horizontal dash (but note that the writers of 182 and 764 seem to use two ticks, one above the other, perhaps representing two asses (see Appendix, 182) [drawing]; in the we described this symbol as having "a short vertical descender which is sometimes combined with the following symbol for one sixteenth" but we now think that this always belongs with the symbol for one-sixteenth = 1 as and presumably represents a, the whole symbol to be interpreted as a cursive a with a ligatured i. [drawing]. These symbols can be seen on the Plate and they can now be more securely identified in the other Vindolanda texts in which they occur (182, 193 etc.). We can also now identify a text in which a is used as an abbreviation for as, as it is in P.Oxy. IV 737 = CPL 311 = ChLA V 308, namely 192.5, where we should read: p(ondo) i a(sses) xi s(emissem) f(iunt) [. It should be noted that the as can also be represented by a horizontal tick which can be used singly and multiply; evidently both conventions were used, sometimes in the same account (e.g. 182). In the present text we have resolved these symbols by using the Latin terms quadrans, octans and as i, but we make no claim to grammatical exactitude; they should be taken merely as indicating what we think they represent.
As for other abbreviations, it is worth noting the use of m in a context which rules out the common m(odius), sometimes but not always with a superscript bar (lines ii.12, 13, 19, 20, 21 and 22, see ii.12-13 note). Also noteworthy is the abbreviation s which we resolve as s(ingularis); this has not hitherto occurred at Vindolanda.
The text is written in a squarish, competent hand with a marked use of serifs, especially on i, l, n and u. Note one or two occurrences of a in the second form illustrated in Tab.Vindol.II, p.53. When s is used as an abbreviation for singularis (but not when it represents semis) it is preceded and followed by medial points (see Adams (2003)).
For the sake of clarity of presentation, we have numbered the lines individually in this text.


]ḷum () xiix · () [
] . ṛiclum () xx .. · () viii . [ n
..riclum traces n
.ariṇ .. () xxi traces n
5 .ar… () x̣x̣x̣ traces n
() lx̣iii [
traces of 2 lines
traces co………. n
10 traces


'Necklace-locks (?), number 2, 3 5/8+ denarii each, total 7 1/4 + denarii.
Cloaks, number 6, 11 1/2 denarii each, total 69 denarii.
Headbands, number 5, 3/4 denarius each, total 3 3/4 denarii.
Hair, 9 pounds in weight, 5 3/4 denarii per pound, total 51 3/4 denarii.
Drawers, number 10, 2 1/2 denarii each, total 25 denarii.
Saddle (?), number 1, 12 denarii.
[[Cloaks made of bark (?), number 15 (?), ]]
Cloaks made of bark (?), number 15, measure(s) (?) .., 3 denarii per measure
(?), total 236+ denarii.
Bags, number 10, 5/8 denarius and 1 as each, total 6 7/8 denarii.
Bowls, number 4, 5 denarii 1 as each, total 20 1/4 denarii.
Bowls, number 4, 3 7/8 denarii and 1 as each, total 15 3/4 denarii.
Bowls, number 4, 2 5/8 denarii and 1 as each, total 10 3/4 denarii.
Reins, number 2, 3 1/2 denarii each, total 7 denarii.
Scarlet curtain (?), 1, measuring 11 1/2, total 54 5/8 denarii.
Greenish-yellow curtain (?), 1, measuring 11 1/2, total 46 3/4 denarii.
Purple curtains (?), 2, measuring 11 1/2, total 99 5/8 denarii.
curtain (?), 1, measuring 10 1/2, total 55 1/8 denarii. '


i.1-2. The tablet is broken at the top left corner and only the feet of the first few letters remain. The traces are compatible with c̣ḷ[a]ụsuras, which we suggest as a possible reading with some confidence; our main misgiving about this is that the first would seem to be somewhat larger than we would expect and it is possible that we should read instead c̣ḷ[a]ụssuras. For the cf. lumbaria in line 9. We note the clusura alexandrina in P.Freib.2 = ChLA XI 485, where the editors suggest that it means the clasp of a necklace. In CIL II 3386 the word seems to refer to a necklace in an Isiac dedication. The price in our account might suggest a complete necklace, something more substantial and valuable than a clasp or lock unless it were made of silver. Alternatively, it may simply mean a lock, as at Vulgate, Bar. 6.17. Note also CIL XII 1121, arcum cum ostiis et clu[suris]. The superscript bar used as a mark of abbreviation over n is wavy in contrast to the flat dash which is usual in the Vindolanda texts. This half of the diptych is also broken at the top right corner and some of the digits have been lost. Our restoration of line 2 is based on the calculation that the objects cost 3 1/2+ denarii each and the total is 7+ denarii. The trace of the symbol after s(emis) looks to be quite a good octans; we have therefore restored 7 1/4 denarii in line 2.

i.3. infiblatoria: the reading is certain and the word has not appeared before in Latin. The only testimony for it appears to be a Greek transcription in a letter of Trajan (c. AD 102) from Pessinus in Phrygia (Rev.Philol. 11 (1937), 106, no.II.7), [fi]mfeiblat≈ria dÊo in a very damaged context apparently referring to a gift of two cloaks (we are grateful to Dr. J.-P.Wild for this reference). It is interesting that this reference is from exactly the same period as the Vindolanda account. The word fibulatorium (sc. sagum) can mean a garment that is fastened, as at Ed.Diocl.19.24, HA, Tyr.Trig.10.12 (cf. CGL IV 361.10, IGRR III.228, P.Oxy.VII 1051.6, 11), i.e. a cloak pinned at the right shoulder with a fibula; for infibulo meaning "to fasten with a brooch (or buckle)", see TLL VII 1410, citing Cassiodorus, Ios.Antiq.3.11, p.75, superhumerale ... sardonychi lapides infibulant. For the use of fibulae as fastenings for cloaks see Wild (1985), 395-6. On the syncope see Adams (2003).

i.5. capitularia: the plural of capitulare rather than capitularium, which means "poll-tax". The word refers to headgear, see Isidorus, Orig.19.31.3, capitulum est, quod uulgo capitulare dicunt, idem et cappa. uel quod duos apices ut cappa littera habeat uel quia capitis ornamentum est. It seems unlikely that the word could denote the moss cap from Vindolanda (see VRR III, 84 and Plate XIII), which is difficult to interpret. As Tomlin points out (Tab.Sulis 55.5 note), the word is glossed in Greek as kefalÒdesmow (CGL II 369.63), "headband", a term which is used in the Greek version of Ed.Diocl. 26.24 to translate capicularium. The capitulare appears at Bath (Tab.Sulis 55.5) and at Caistor-by-Norwich (Britannia 13 (1982), 408, no.9, cap[t]olare). For evidence for headgear from the northwestern provinces see Wild (1985), 383 (male), 392, 395, 401-2 (female), and (1968), 177-8, 186-7.

i.7-8. cap̣illamentị : this can be used of the foliage of plants (see TLL III 312), but it is more likely to refer to hair, either human or animal. The hair is calculated by weight and the price seems high, at 5 3/4 denarii per pound, but on the other hand one would get a lot of hair to the pound. We have thought of several possible uses to which hair might be put but we can suggest no way of deciding which is more likely. (i) Given the occurrence of other items of equestrian equipment in this account (lines 10, 18), it is possible that horsehair was used for stuffing saddles: see Connolly (1987), 12, Bishop (1988), 104, van Driel-Murray (1989), 281-318, with discussion of the saddle remains from Vindolanda (cf. VRR III, 17-8), and that from Newstead which may, however, have been stuffed with chaff (VRR III, 311); for saddles stuffed with goose-feathers see Fronto, 2.1 (Van den Hout ed., p.122), a sedilibus equitum pluma quasi anseribus deuolsa. (ii) Hair ropes for catapult springs were made of both animal hair and female hair, see Marsden (1969), 17, 53, 83, 87-8, and cf. HA, Max. 33.1, ut funes de capillis muliebribus facerent. (iii) Women's hair was also used for wigs (Ovid, Ars amatoria 3.165ff.), the hair of German women being particularly admired, cf. McKeown (1989), 381.

i.9. lumbaria: TLL VII 1805, lumbare (also lumbarium, TLL VII 1806). This is a garment worn around the lower part of the torso, glossed as subligar (cf. 346.i.4) at CGL V 603.21; see also P.Freib.2 = ChLA XI 485, where the editors suggest that it means a "ceinture ou boucle de ceinture", CIL VIII 4508 (Zarai tariff). For discussion of "drawers" see Wild (1985), 382. Gladiators wore a loincloth which might offer some guidance, but the Latin name for that garment is not attested.

i.10. scordiscum: the meaning is problematic; see Adams (1992b), 159-68, arguing that in some contexts it must mean a leather covering put on a horse's back, rather than a saddle, but admitting the possibility (167) that it might sometimes also designate a saddle, especially at Ed.Diocl. 10.2 and Not.Dig. IX.18, p.145 (Seeck ed.). The present context does not allow us to decide one way or the other. At Ed.Diocl. 10.2 the item costs 500 denarii; the price of 12 denarii in the Vindolanda account is perhaps more suggestive of a saddle than a saddle-cloth.

i.11. This entry was not completed and is crossed out. It seems to be corrected in the following line. The writer may have either written 15 1/2 by mistake or written s for singulare, in which case there will have been medial dots on either side of the letter which are now obliterated by the crossing-out. It may be that the whole line was crossed out because the writer realised that he did not have the space to include the whole of a two-line entry at the bottom of the left-hand side of the diptych.

ii.12-13. Despite the loss of most of the first word in line 12, it seems obvious that the second word is the same as in the previous line and what remains of the first word is consistent with sa]g̣a, suggesting a repetition of the aborted entry in i.11. There are two problems with this entry. First, the reading and meaning of the adjective qualifying saga and second the interpretation of the accounting in the entry as a whole. The reading of the adjective, which occurs in lines 11 and 12, seems to be either porticia or corticia but the latter looks much better in line 12. It seems likely that this is a form of the word corticeus, which means "made of bark" (OLD s.v.). There are references to clothing made of bark but only in contexts which suggest that this is regarded as a primitive practice: Seneca, Ep. 90.16, non corticibus arborum pleraque gentes tegunt corpora, Arnobius, Aduersus nationes 2.66, quid enim si, hoc modo culpam uelimus infligere prioribus illis atque antiquissimis saeculis ... quod corticibus contegi et amiciri desierint pellibus. Dr. Wild remarked (per litt.) that cortex might include the bast fibres on the inside of the bark of trees which were certainly used for very fine textiles in the Swiss Neolithic before flax replaced it. DML has a 14th-century reference to corticium possibly meaning a kind of garment, but this seems unlikely to be relevant. TLL IV 1071, s.v. corticatus, refers to its occurrence in CGL III 370.24 where it is glossed by fulÊrinon [sic = filÊrinon]. This is to be compared with CGL III 22.19, where the MSS have fuliron porticatum [sic] which is in a section headed de coloribus. The idea of "lime-coloured cloaks", however, is not attractive. The interpretation of the accounting is complicated by the loss at the ends of lines 12 and 13 and our suggestions must therefore be regarded as hypothetical. The beginning of line 13 is also lost but this could well have been inset, so that nothing may be missing here. The abbreviation m also occurs in lines 19-22, but there is no guarantee that it stands for the same word there; note that here the abbreviation m has a medial dot before and after it whereas in lines 19-22 it is written with a superscript dash. Elsewhere in accounts in the tablets this is to be resolved as m(odius), which the context rules out here. We think it possible that it stands for either modus (cf. OLD s.v.3a), modulus (cf. TLL VIII 1249ff.) or mensura (contrasted with pondera in OLD s.v.5a). In all cases the meaning would be the same, a unit of measurement for textiles; and we should need to suppose that there was some conventional measure in use which was well enough known not to need further specification. The entry would then take the form: saga, quantity 15: units of measurement nn; for each unit of measurement 3 denarii; total 236+ denarii. Supplementing the total to the nearest figure which is divisible by 3, for the sake of simplicity, gives us 237 denarii, which will be the cost of 79 units of measurement. We suppose that the accounting is on the basis not of the cost of each sagum, but of each unit of measurement, whatever that is. If 15 saga yield 79 units of measurement, the average for each sagum will be just over 5 units and the average cost just over 15 denarii (this seems very expensive, cf. 184.20); the cloaks may well, of course, be of different sizes for men, women and children. They will have been woven individually and not cut from larger bolts. It would not be unreasonable to suppose that the unit was in fact one foot or thereabouts (see the illustrations in Wild (1985)).

ii.14. sarcina is normally baggage, but here we presume that these inexpensive items are bags, perhaps for use with pack-animals; see the evidence collected by Lepper and Frere (1988), 268, Adams (1993), 58, and (1995b), 122.

ii.15-17. In the we translated trulla as "skillet", adding that it is the common, handled skillet, of which several examples have been found at Vindolanda. RIB II 2415 contains inscriptions on such vessels to which the editor assigns the term trullae, but we know of no example on which the word itself is inscribed (cf. in particular 2415.39 with Boon (1984), 403-7). The testimonia assembled by Hilgers (1969), 364 show a variety of uses for these vessels: as ladles or scoops (cf. OLD s.v.), cooking pots, dishes or bowls and cups for eating and drinking, which could clearly be quite expensive and valuable. We see no way of specifying what kind of trulla is indicated in the present text and we have reverted to the translation "bowl" which we used in 194.B6 (and see note); the differences in price here presumably indicate vessels of different size or quality. See also contrullium in 588.i.3 and note.

ii.17. We now think the first figure is better read as 2 1/2 1/8 denarii 1 as and the second as 10 1/2 _ denarii.

ii.18. frẹnos: the noun is usually neuter in the singular and masculine in the plural, see OLD s.v.; the form is not significant, see Adams (2003). For leather strapware from Vindolanda see VRR III, 18-9 and Plate X.

ii.19-22. For uelum = "woven cloth" see e.g. Pliny, NH 35.150. It can, however, be something more specific, such as "curtain" (see OLD s.v.4), an interpretation favoured by Dr.Wild. These will then be dyed wool curtains. It is odd, however, that there are three single curtains and only one pair. It is also noteworthy that the numeral is in the form of a long i with a medial dot on either side, whereas in line 10 we have n(umero) plus a normal sized i. On the interpretation of the abbreviation m as a unit of measurement see note to line ii.12-13. If this is correct, the cloths all come in lengths of 11 1/2 units except for the last one which is 10 1/2. Note that in the we did not offer readings for the small fractions at the end of each of these lines.

ii.19. coccini(um): technically this means "kermes-dyed", i.e bright scarlet, especially since it is here distinguished from purpureum, see ii.21 note.

ii.20. uirdem: there is no doubt about the reading, although uirde is what we expect; on the form see Adams (2003). On the colour cf. O.Max. inv.254, lines 3-4, fasciclos duos uirdiorum (Bülow-Jacobsen, Cuvigny, Fournet (2003), no. IV, 35). "uiride is a fairly rare textile colour in antiquity, achieved with a combination of yellow and blue. Since both weld for yellow and woad for blue are attested in Roman Britain, a green curtain need not be exotic or very expensive - just a bit unusual" (J.-P.Wild per litt.).

ii.21. purp(urea): "this need not mean murex-purple, which would be very expensive indeed, but a combination - woad (for blue) overdyed with madder (red). Madder and/or a bedstraw has been positively identified in the Vindolanda textiles and there is evidence, too, for a purple won from a lichen (Taylor (1983), 115-24)" (J.-P.Wild pers comm.). In the we read the numeral as lxxxiix but on the scan it looks much more like lxxxxix.

ii.22. uelum ......: the writing after uelum is obscured by dirt and very unclear. The initial letter is c, g or p, followed by a or r; later there is a circle which looks like o, but which could be the bottom of d or even b with the top lost. We presumably need a colour word and we have considered candidus, galbinus and prasinus, but do not find any of them wholly convincing; gạḷḅin with the end of the word abbreviated (as in lines 19 and 21), is perhaps the least unlikely.

Back. This appears to be a similar text but we cannot establish an exact correspondence with the format of the text on the front. Each entry seems to begin with a commodity but, unlike the text on the front, the nouns appear to be in the singular. There then seems to be a sum of denarii, quite large in the cases where we can make out the digits. In some lines there are then abraded marks which might correspond to the s(ingularis) on the front, and there appear to be further denarius symbols and figures at the right, but these do not suggest a multiplication of the first figures given in the entry. Line 6 seems to be an overrun of the entry beginning in line 5 and the same may be true of lines 7 and 8.

Back.2-3. We seem to have the same noun in these two lines (perhaps also in line 1 although not enough survives to be sure). Perhaps u]ẹṛiclum, a javelin, or (e)u]ẹṛ(r)iclum, a net (cf. 593.i.4). Of the other possibilities offered by LVL, galericlum is too long; fericlum (cf. OLD s.v. ferculum), "tray", is also plausible.

Back.4-5. Perhaps the same noun repeated. The remains in 4 suggest .ariṇ.. and we can imagine far in 5; if we do have the same word repeated this suggests f̣ariṇạṃ.

Back.9. This is anomalous in that the word is set out to the right, directly below the symbols and digits in the line above, so the entry may follow a different pattern. The traces surviving at the right look like co...ịṇ..ụṃ; their position suggests that there is no reason to assume that co is the beginning of a word.

Download EpiDoc version using the CC license Creative Commons License and EpiDoc Schema v.5