Publication: Bowman-Thomas, VWT p.28
References: Bowman, Britannia v (1974), p.366; Bowman-Thomas-Wright, Britannia v (1974), pp. 477-80, pl. XLIII; Wright, Proc.XIVth Int. Congress of Papyrologists, p.355; J.N. Adams, BICS 22 (1975), pp.21-3; Birley, Vindolanda, p.155; Bowman, ZPE 18 (1975), p.239, n.8; Bowman-Thomas, Historia 24 (1975), p.476
This document consists of a diptych which was scored and folded with the writing on the inner faces. The outer faces are blank. Comparison of the dimensions of the surviving pieces with other complete examples suggests that the loss at the top and bottom might be about 1.5 cm at most and at the right-hand side perhaps 2 cm. The loss of the top and bottom edges will have removed the tie-holes, if such there were, and it is thus impossible to be sure whether this document consisted of a single diptych or a series tied together, as in No. 4 (190) (see Vol. 1., Ch. 2). Like No. 4 (190), this text is written across the grain of the wood and parallel with the short edge of the tablet. It will have had to be opened longitudinally, thus presenting a single continuous column. The habit of using tablets in this way for accounts is paralleled in a very similar text on a wax tablet from Dacia (CIL 32, p.953 No. XV, pag.posterior), which closely resembles the Vindolanda text in several ways. It begins with a date-heading (we cannot tell whether No. 5 (191) began similarly because the top is lost, but it is not at all unlikely) and continues with a list of items of food followed by amounts in denarii. The items are in the accusative case, except for two which are in the genitive, and are followed by the sextarius symbol (f). This does not appear in the present Vindolanda account but it is to be found in No. 4 (151). It is clear that in the Dacian tablets the sextarius symbol explains the use of the genitive. In No. 5 (191)the genitive is used in line 5 and probably in line 16, whereas we have the accusative in lines 4, 6 and 7, all meat items; the other case-endings are lost. In the editio princeps we restored all other endings of words for meat in the accusative, but used the genitive (understanding a word of quantity) for the non-meat items. We added that this was ‘somewhat arbitrary’. J. N. Adams, in the article cited above, examined the language of this document and stressed that alternation between accusative and genitive in such lists may be entirely haphazard. This is correct, but the evidence of the Dacian tablet supports the use of accusative and genitive in the manner assumed in the editio princeps, and we have therefore retained the same restorations of the case-endings in the present edition. Adams supposed that the meat items must have been followed by a weight (and cited examples of this), but the Dacian text shows that this is not necessarily the case: in line 16 it has porcellum (den.) v. We might suppose that similarly in the Vindolanda text all the meat items were followed directly by an amount in denarii, that is, it recorded the purchase of ‘meat at a price of n denarii’. Unfortunately, in the only line where any trace of ink survives after the meat item (line 7) we have been unable to read what follows pernam.
There is no reason to believe that No. 5 (191)is part of the same series of accounts as No. 4 (190). The hands in which they are written are markedly dissimilar. The present writer uses a right-sloping, not inelegant form of Latin cursive with tall, slim letter forms. Note the large initial p in line 6, the shape of n with the cross-stroke almost horizontal, and the two forms of d (notably in line 12). Although both Nos. 4 (190)and 5 (191)deal with food, no item which is in one recurs in the other. It seems likely that the items in No. 5 (191) represent purchases made on behalf of one of the units stationed at Vindolanda. The most noteworthy feature is the high proportion of meat and its variety. Until very recently scholars, perhaps misled by Vegetius, were in the habit of saying that Roman soldiers ate little meat (e.g. Webster, RIA2 p.255, ‘meat was certainly eaten on special occasions, festivals and feasts but was not regarded as a daily need’). R. W. Davies has suggested (Britannia ii (1971), pp. 122-42), on the basis of a study of bone remains in Roman forts, that this is not true and that soldiers consumed a good deal of meat. The animal remains found at Vindolanda certainly fit this picture (see Birley, Vindolanda, pp.122-3). The present text would seem to confirm this (all the meats mentioned, except perhaps venison, were reasonably common), but a word of caution is necessary. There is no means of knowing for certain whether this list represents the soldiers’ regular diet or whether it refers, for example, to officers’ rations only or to special commodities purchased for a feast. The Dacian account contains both meat and non-meat items; it includes only porcellum of the meats in No. 5 (191) and only sal of the other items (but it does have acetum, which is in No. 4.31 (190)). Its editor suggested that the account might refer to the Larales (but note that there is no evident connection in the text with the army) and part of the account in No. 4 might well relate to the festival of Fors Fortuna (see No. 4 introductory comments (190)). On the other hand, the reading of line 12 in No. 5 (191), if correct, suggests that there is a good chance that at least part of this account records purchases for daily use (see note ad loc.). For more general matters relating to diet, stores and provisions and the bookkeeping connected with them see Vol. 1., Ch. 3.
1. in p[: the reading is probably the same as line 8 and line 11 (see note ad loc.).
2. ]s: there are several dark marks visible on the plate at this point which seem not to be ink. In the editio princeps we read ]c but we are now satisfied that it is just as easy to read ]s. s, which would presumably stand for s(umma), cf. line 15, makes more sense. It is perhaps worth noting that the Dacian account (see introductory comments) has a date-heading followed by a total credit from which individual sums are then deducted. Possibly our text began with some such entry as acc(epti) denarii followed by a numeral.
3. condiment[orum: for the restoration of the genitive see introductory comments. Compare Ed.Diocl. 6.48, condimen[torum] praemis quorum fa<sce>s n. octo. On this and the items referred to below in Diocletian's Price Edict there are excellent bibliographies in the notes of Lauffer, DP, ad locc.
4. capream: also in line 13, the only item to occur twice. In the editio princeps we took this to mean goat’s meat and supposed it equivalent to capra. In the fourth chapter of Diocletian's Price Edict the meats listed include all the others which occur in our text and Ed.Diocl. 4.45 refers to capra, which is certainly goat's meat because of the Greek equivalent . However, TLL and OLD distinguish capra and caprea and regard the latter as the name for roe-deer. It might be felt that a reference in our text to roe-deer is unlikely because of the occurrence in line 10 of ceruina (venison). But it is not at all unlikely that by ceruina was meant the flesh of red deer (deer par excellence) . We know that both red and roe-deer existed in the vicinity of Vindolanda in Roman times (but not fallow deer which were not introduced until the thirteenth century), and that (in the middle ages and later) roe-deer were regarded as inferior; so the term ‘venison’ could be restricted to the flesh of red or fallow deer, cf. OED s.v. (we are indebted for some of this information to Mrs. J. L. Drury).
5. salis .[: just before the break at the edge there is a trace of ink which could be part of a superscript bar as used over m (for modii) in No. 4 (190) (see note to line 5). But we cannot be sure whether m or another kind of measure might have been written (but see Cicero, Amic.67, multos modios salis simul edendos esse, ut amicitiae munus expletum sit). The weight abbreviations p, q would be appropriate for a commodity like salt. Dr R.W. Davies pointed out to us that salt might have been used for preserving the meat. It also appears in CIL 32 p.953 no.XV.25, salem et cep(am).
6. porcellum: cf. CIL 32 p.953 no.XV.16, Ed.Diocl. 4.46, porcelli lanctantis (sic).
7. pernam: cf. Ed.Diocl. 4.8, per<n>ae optimae petasonis.
8. in p[: see note to line 11.
10. ceruin[am: used as a short form of ceruina caro, see TLL s.v., cf. Ed.Diocl. 4.44, cerbinae; see J.N. Adams, BICS 22 (1975), p.23. The fold in the tablet comes between this and the next line.
11. in p.[: this reading ignores an apparent vertical stroke between in and p. Although the top of the first letter may be lost and the first part of n could well be a it does not seem possible to read a form of far or farina. We are confident that the apparent traces before i are not ink and that the line is indented like lines 1 and 8 and contains the same entry. In this line p is certain and in the other lines it is possible if not probable. In Ed.Diocl. 4 the meat is measured by weight, usually abbreviated po(ndo); this is not normally preceded by in but in po(ndo) does occur in Ed.Diocl. 4.46-8, 7.24-8, perhaps cf. also Pliny, NH 33.158, pretium in pondo libras denarii duo. Other possibilities which have occurred to us are in pretio (precio), a regular expression in mediaeval accounts, and in praetorio or in praetorium (cf. perhaps RMR 82.7). In this line, just as the tablet breaks off, there may well be traces of a letter (or symbol), but they are too indecisive to help us solve the problem.
12. ad cotidia[n: a good, but not certain, reading. Further cleaning since the photograph was taken proves o certain; the foot of the first i has been fouled by the top of c from the line below. The apparent down-stroke after d is merely the grain of the wood. The word also occurs in No. 44.3 (521) but that is probably a letter. Its implication in the context of No. 5 (191) would be that the items under this heading were for daily use, whereas the earlier entries may fall into some other category (see introductory comments and note to line 11). This, of course, does not preclude the supposition that all the entries in this account may have been grouped under one (lost) date-heading, in contrast to the series of date-headings recorded in No. 4 (190).
14. [[s X[: ‘[[..s[’, (editio princeps). We are now confident that s was followed by the denarius sign. The line was crossed through several times by the writer, presumably because he made a mistake in his calculation (cf. line 15).
15. The interpretation of s as s(umma) seems inescapable in the context. For the abbreviation cf. RMR 69.8 etc. We think it distinctly possible that this line (and therefore no doubt the deletion of the previous line) was the work of a second hand.
16. bracis: the writing commences further to the left than in the preceding three lines, almost directly under ad in line 12; what precedes it is almost certainly not ink. The writing is clear as far as brac; in the editio princeps we were uncertain whether to read es or is after this, but we now feel reasonably confident that the correct reading is bracis. Lewis and Short record a word brace (genitive braces) but seem to base this solely on the Pliny passage quoted below, where the word appears in its accusative form as bracem. We follow TLL and OLD where the nominative is considered to be bracis (doubtfully in OLD). In that case bracis in our text is likely to be genitive singular (we think accusative plural less likely). On this Celtic loan word see J.N. Adams, BICS 22 (1975), pp.22-3. The word suits the context very well. Pliny, NH 18.62 says of it: Galliae quoque suum genus farris dedere quod illic bracem vocant. Corpus Gloss. Lat.v.616.26 has: braces sunt unde fit ceruisia (cf. No. 4.12 note (190)). It is therefore a Gallic name for a kind of cereal which could be used for malting (see N. Jásny, Wheats in Classical Antiquity (Johns Hopkins Studies in Historical and Political Science 42, 1944), pp.134-5). A. Holder, AS, s.v. braci- adds a reference to Ionae uita Columbani c.21, ed. Sur. 21 nov., p.475: centum dicens modia esse uini, ducentaque frumenti, sed et bracis centum modia, thus distinguishing bracis from frumentum.
18. ].um[: before m we appear to have u, although u is not made quite like this elsewhere in this hand (n is not wholly impossible); before it the top of a tall letter, c, e, f or s. We cannot read frum[enti as there would be no room for r. In the editio princeps we suggested sum[ma (written in full as indicating the final total for the whole amount?) or sum[en. We still think both these are feasible alternatives, but the curve at the top of the first letter is more like e in ceruin[am (line 10) than c, f or s and it is possible that it was preceded by two or even three letters.