Tablet 643

description (a) 182 x 53 mm (approx.) (b) 19 x 18 mm. Plate 15.
Archaeological data. Location: LXXVIII/W. Period: 3?.
Though made up of several fragments and with some pieces missing, the diptych is mostly complete at the top and both sides, and probably almost complete at the foot (see i.6 note (end), ii.5 note). Its format is unique among the Vindolanda material. On the front we have the start of two letters from a certain Florus; one is addressed to Calavir(us) and the other to Titus. On the back of the one addressed to Calavir(us) is the conclusion of a letter. These are all written in the same hand. In addition we have a closing greeting written at right-angles down the left-hand margin of the letter to Titus; this is probably also in the same hand. There is a small detached fragment (b), which we suggest belongs at the left of lines i.5-6: see the note there.
We are reasonably confident that the letter to Calavir(us) was written in the left-hand column and that to Titus in the right-hand. One argument is that it is much commoner in the Vindolanda tablets for marginal additions to be added between the columns (see 302, 311, 349, 352, 645, 694, 702, 705, 706; only 316 is likely to have the addition at the left of the left-hand column). We also believe that it makes better sense to take the writing on the back as the conclusion of the letter to Titus. In which case Florus wrote the whole of his first letter in the left-hand column, then began his second letter at the top of the right-hand column. He then turned the tablet over and completed the second letter on the back of what is the right-hand column on the front, the place where writers normally put the address only. Florus, however, first completed his letter on this side before adding the address: Caelouiro dabes. This form of address is not attested elsewhere in the Vindolanda material but is a common form in the tablets found at Vindonissa (Tab.Vindon., pp. 36-7) and elsewhere (cf. CEL II, p. 41, but note that it does not occur in CEL 91 = Tab.Vindol.I 23: see the republication as 263). We understand the words to be an instruction to the letter-carrier, who is likely to have been the man mentioned in ii.3 (see ii.4-5 note).
There are a number of examples in papyrus letters of two letters from the same person to two different people being written on a single sheet of papyrus. Good parallels are P.Tebt. II 416 and SB III 6263. The name of the sender of our letters, Florus, is found also in 258, 281 (see Appendix for both), 601 and 644. 258, 281 and 644 are letters from Cluvius Florus, who belonged to the ala Petriana (see 644, introduction). 601 is an account headed ratio Flori. None of these is written by the same hand as the present text and there are no strong reasons for suggesting either man is the Florus who sent the letters (the cognomen is common). If, nevertheless, the Florus in the account is the same as the man in the present text, this would suggest that these letters are mere drafts or file copies, since we should expect the account to relate to a man who was at Vindolanda. The mention of a casula in ii.3 might point to Titus being a civilian; on the other hand we have a reference to a beneficiarius in i.5.
The hand is a rather crude and inelegant cursive: note the form of l which curls to the left at the foot, and b in which the normal closed loop at the bottom is simply a diagonal tick. Word division is generally good. The texts of the letters are noteworthy as being among the poorest compositions from the linguistic point of view, if we have understood them correctly: see the notes to a.i.2, 5, a.ii.2, 3, the consistent use of -es, -et for the endings of the future tense, q (and not qu) for c in a.i.3, and the inconsistency in the spelling of the names in a.i.1 and back 5; note too the odd word order in a.back 3-4 and ii.2. See further Adams (2003), §3 and §6.
The matters dealt with in the letters are not without interest: instructions in the first to give a locked box to a beneficiarius, which he is probably to seal with his signet ring (see a.i.5-6 note); and in the second to give an axe to someone. The name of the recipient of the first letter, variously spelt Calavir(us) and Caelovir(us), is not elsewhere attested. Both letters read as though written by a superior giving instructions to subordinates.


Florus Tito suo ṣạḷutem n
frates securem quam iṇ n n
casula habea dabes G̣ạṃ[.. n n
qui ..[...].ctiliạ[.. n
5 dabet et ut re[[.]]ddat n n
. . . . . . . .
opto benẹ [ n
. . . . . . . .
. . . . . . . .


'Florus to his Calavir(us), greetings. The closed small box and whatever things have been locked in it(?) give to the beneficiarius which(?) he will seal with his ring."
Florus to his Titus, greetings. Brother, if you happen to have(?) an axe in your hut, give it to Gam-, the man who will deliver you this tablet(?), and in order that he gives it back and do not give it to him except on condition that he straightway places it in the cart. Ingenua, your daughter, sends (both of) you greetings. Deliver to Caelovir(us). (Margin) I pray that you are in good health. '


a.i.1. Calauiro: A.R.Birley (2002), 109 records the name as "Caltavirus (or perhaps Ealtavirus)". It is possible to see why the initial letter could be read as e, though we regard c as virtually certain both here and on the back, but there is no justification for the insertion of t. We do not think there is any serious doubt about the correctness of the reading Calauiro here, although it is equally clearly written Caelouiro in a.back 5. We cannot precisely parallel the name in either form. RNGCL and NPEL list several names beginning with Cal-. RNGCL also has examples of names beginning Cael-; NPEL has one example from Narbonensis of an incomplete name beginning Caelo-. For variant spellings of a name in one and the same letter see 310 (Veldeius, Veldedeius).

a.i.2. The layout of the address does not follow the common pattern in which salutem is written in the second line, justified at the right, with the letter proper beginning on the third line. Although one or two other texts differ slightly from this pattern, the present text seems to be the only example where the letter proper begins on the same line as salutem. clusa: the writing runs up to the edge of this section of the diptych. If we have a left-hand column (as we think most probable), we should have expected to see part of any further letter on the right-hand column. However, no trace remains at this point and we are reasonably certain that final m, which is necessary for grammar and sense, i.e. clusam, was not written; on this phenomenon see Adams (1995a), 88.

a.i.3. quequmque: for quaecumque; the third q may have been corrected. Adams (1995a), 87-8, notes that in the tablets the digraph ae is written with remarkable consistency.

a.i.3-4. At the start of line 4 is almost certain with only one letter lost before it. At the end of line 3 there may be a trace of ink at the edge, but the tablet is blank before this and we do not think this trace is writing. inị is a good reading but is difficult to explain; hardly simply a mistake for in, with e]ạ in the next line. If we read in ị[l|l]ạ; it is difficult to say why no trace of l survives after ini. Perhaps, as Rea suggests, we should read in ị|[l]ạ and take it as a misspelling of in illa; for this spelling in inscriptions see TLL VII.1 340.37ff.

a.i.4. coṃclusae: it is very hard to read the third letter, of which only the top survives, as n, although the writer must have intended conclusae. For examples of the use of m instead of n see ILS Index XVI, p. 825 (e.g. numcupata). dabes: the form is consistent in this text (a.i.4, a.ii.3, back 5; also dabet ii.5, signabet i.6); see Adams (2003), §3.

a.i.5-6. We suggest that the detached fragment (b) belongs at the start of these lines.

a.i.5. Presumably the first word in this line was the name of the beneficiarius. It may end -ṇo or -ṭo, less probably -ịo. If we are right to fit (b) at this point, it may be just possible to read Armato, but the first two letters are particularly hazardous. benifeciạrio: on the spelling cf. Adams (2003), §3. The title occurs in 180.18 where the name begins Lu- but does not survive in full, and in 344.i.10, where no name is given; cf. also 642.i.8 note.

a.i.6. signabet anulo: if (b) belongs here, the reading at the beginning of the line is either qua[m or qua[s. It was quite usual to seal letters (TLL II, 195.64ff.) and there are surviving examples of sealed stilus tablets (see Marichal (1992), 165-85, especially figs. 7, 8 and 10; RIB II 2411, introduction). We know of no example of the use of a seal with leaf tablets either from Vindolanda or Carlisle. It is more likely, however, that the writer is here concerned about the security of the box and its contents and is asking that the beneficiarius should seal the box with his signet ring as a precaution against theft. This suggests that the contents of the box were important or valuable. In this connection we note the lead seal of a beneficiarius consularis from Piercebridge (Britannia 22 (1991), p.302-3, no.33 = AE 1991.1143); for seals of beneficiarii from Britain see also RIB II 2411.246-7. Lead seals may normally have been made from special dies. For signet rings with retrograde inscriptions see, e.g., RIB II 2422.25, 42. Henig (1975), 19 mentions a silver signet ring from Vindolanda. See also Nelis-Clément (2000), 258 and 365, no. 154a, with further references. There may well be slight traces of a line 7, though the sense is complete without anything further. The writer may have added a greeting to Calavir(us).

a.ii.1. Tito: for the use of praenominal forms as cognomina see 156.2 note. There is no other example of the name Titus in the tablets. For salutem written on the same line as the names of the sender and addressee see, e.g., 212.

a.ii.2. frates: the reading is absolutely clear. We can hardly suppose a slip for fratres, since the letter is addressed to one person only (and cf. dabes in 3). Presumably the writer intended frater. For the word order cf. Tab.Vindon. 52. securem: this form of the accusative is quite well attested: see LS s.v.

a.ii.2-3. The reading habea is clear. We suppose that the writer omitted the final letter (of which there is no trace). There would be no problem in supposing omission of final -m (see i.2 note above), but we cannot then explain the subjunctive. If we suppose the writer intended habeas, a possible explanation of the subjunctive would be that quam is for si quam, leading to the meaning suggested in the translation. We owe this explanation to Adams, who discusses the point further in Adams (2003), §4.

a.ii.3. casula: the word may suggest a modest dwelling, a cottage, which does not seem appropriate to military quarters. This would support the suggestion in the introduction that we may be concerned in these letters with civilians. Alternatively, and perhaps more likely, the word may mean "hut" or "shed". G̣ạṃ[: the reading is not certain (but more probable than G̣ạṇ[). A name seems to be required; Gambax occurs in 184.5 and Gami.oni in 608.2, but both look too long for the space available. Another possibility is G̣ṇạ[, e.g. G̣ṇạ[to (Gnavorix, attested in Appendix, 353.back, is much too long).

a.ii.4-5. The positioning of the fragment which contains the start of 4 and the whole of 5 is guaranteed by b in benẹ (written in the margin): this survives almost entirely on this fragment but the downstroke starts on the main fragment containing lines 1-3. ].ctiliạ[: before ctiliạ the vertical is most probably part of n; also possible is ].i. There is a slight gap between c and t. If this indicates word division, we could easily read ha]ṇc tiliạ[m, to be taken as the object of dabet in line 5. If we are not looking for word division, possibilities include fictilia ("earthenware dishes") and pictilia, which would have to be understood as neuter plural, meaning embroidered or painted things, cf. Apuleius, Met.10.18. If we read ha]ṇc tiliạ[m, before it we may have no more than tibi; for the sense we envisage see the translation. Since wooden leaf tablets would in most parts of the Empire have been made of lime-wood (see Tab.Vindol.I, p. 37), we suggest that tilia may have been the correct term in Latin to describe this type of writing-tablet; cf. OLD s.v. b and Introduction, p. 00; see also Appendix, 259, 589.a.1, and 707. If this is correct, we suppose that the man named at the end of line 3 is the bearer of the letter, to whom Titus is to give the axe.

a.ii.5. There is probably a trace of an ascender from line 6, see also the next note.

Back.1. ẹị ḍaḅ[e]ṣ: all the dotted letters are doubtful so that the reading as a whole is far from secure. Nonetheless this is the sense we expect.

Back.2. carrulo: this diminutive is also found in 315 and 316. ponạ[t]: only a trace of the foot of remains. It might be possible to restore ponạ[s], but in that case we should have expected to see something of the foot of s.

Back.3. Ingenua: the name also occurs in 642.ii.9, where see the note.

Back.3-4. The word order is unusual and we cannot parallel it. Perhaps the writer added u[e]stra filiạ as an afterthought, realising there was more than one Ingenua known to the recipients or filia may be used loosely to describe a young girl known to the recipients. The switch to the plural is also noteworthy (cf. Adams (2003), 00). The words seem to imply that Titus was living at Vindolanda with his wife, an additional reason for thinking he is likely to have been a civilian; but filia need not necessarily be taken literally (see above) or the letter may be a draft.

Back.5. Caelouiro: see a.i.1 note. dabeṣ: in both the papyrus letters referred to in the introduction (P.Tebt. II 416 and SB III 6263) the address on the back begins with épÒdow, followed by the name of just one of the two addressees.

Margin. If after benẹ we supply ualeas in the same line, at least two lines are missing after line 5 of the letter. But there is a trace below n which could be the top of l, i.e. ua]ḷ[eas could have been written on a second line. In that case no more than one line need have been lost after line 5.

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