This diptych, containing a letter written in two columns, is preserved complete. The tablet is extremely thin and should be compared for this and other features with No. 30 (295) (see introductory comments there). The folding in this case is of particular significance because, when the tablet was found, the two halves of the diptych were still firmly joined together. There are two V-shaped notches in the left and right-hand edges (see Vol. 1., Ch. 2) but no sign of any tie-holes. The writing on the inner faces is in two columns. The left-hand column is broader than the right (cf. Nos. 22 (250), 30 (295)) and extends beyond the central fold (which was presumably made after the letter was written). The central area on both sides of the fold is defaced by offsets (cf. No. 39 (299)) which suggests that the writer folded the letter while the ink was still wet.
The letter is addressed to Cerialis by Niger and Brocchus. The latter does not appear elsewhere in the Vindolanda tablets but it may be that Niger is to be identified with Oppius (?) Niger, who was perhaps the commander of the unit at Bremetennacum (Ribchester), see No. 30 (295) introductory comments and line 6 note. In general, the main hand of No. 21 (248) may be described as regular and elegant; the pen-strokes are fine and the letters carefully made. There is a tendency to enlarge the first letter of each line and to exaggerate ascenders and descenders. Perhaps the most noteworthy feature of the letter-forms is their variety: contrast the tall 1 in salutem (line 2) with the short form in felicissimum (line 4), long i in precari (line 8) with the short form prevalent elsewhere, and the marked difference in the size of s and the tail of r. There is no use of interpunct and word division is only occasionally marked. The writing slopes to the right and is very sparing in its use of ligature, which is scarcely found except in us (lines 3, 9, 11, contrast line 4). In format and in the general style of the writing this tablet closely resembles No. 30 (295). Whether both are in the same hand and thus from the same Niger (though to different addressees) is further considered in the introduction to No. 30 (295).
The letter contains an expression of the hopes of the writers for the success of something which Cerialis is about to do. We cannot be sure whether this relates to a specific enterprise (e.g. a military campaign) or is more general. But the final sentence of the main text, which suggests that Cerialis will soon meet the governor, perhaps slightly favours the first notion. There might well have been opportunities for military success in the period when the Romans were engaged in the withdrawal from southern Scotland.
1. Niger: he might be the same person as the writer of No. 30, Oppius (?) Niger (see No. 30 (295) introductory comments and line 6 note); the tone and content of both letters suggest that he was of similar status to the addressees, both of whom held the post of praefectus cohortis.
Brocchus: the name appears only here in the Vindolanda tablets. It is a fairly well-attested cognomen, see Kajanto, LC p.238. Its origin is perhaps Italic rather than Celtic (see J.S. Conway, J. Whatmough, S.E. Johnson, Pre-Italic Dialects of Italy (1933), viiic, xic, xiia, xxivc; D. Ellis Evans, Gaulish Personal Names (1967), p. l59 n.5), though A. Holder, AS s.v. says it is part Celtic, part Roman.
Ceriali: the reading of the name is certain here despite the loss of the tops of letters at the end. We see no objection to the supposition that the same Niger wrote to both Cerialis and Crispinus (No. 30 (295)) at Vindolanda and that they were prefects of cohorts which garrisoned the fort in succession during the period A.D.95-105 (see Vol. I., Ch. 3).
2. suo salutem: this format, with both words on the second line and at the right-hand side, is unique in the Vindolanda letters.
3. Optamus: there appears to be a slanting ink-stroke above the o (o and p are noticeably enlarged). We think this is probably unintentional. It has occurred to us that it might be an apex mark; but (a) there is no consistent use of this mark in the text (the only other possible example is over to in line 8, see note), and (b) this is normally used only to mark long vowels: since its original purpose would have been to indicate quantity in cases where it was ambiguous, there would be no reason to place it over a short vowel lengthened by two following consonants.
it quot: the substitution of final t for d and vice versa is very common in vulgar Latin, see P.Mich.467.24 (=CPL 250), 468.2 (=CPL 251), 471.13 (=CPL 254) and cf. Adams, VLLCT pp.25-9.
5. autem: for its force here see OLD s.v.4.
6-7. quom . . . conueniat: for quom see Vol. 1., Ch. 5. We have not been able to parallel the expression precisely. For this sense of conuenio followed by the infinitive see TLL s.v.III.B.lb (cols.834-5) and cf. Pliny Ep.2.6.6, conuenit autem amori in te meo . . . sub exemplo praemonere quid debeas fugere. For uotis we might compare No. 37.8-9 (225), inter praecipua uoti habeo (see note ad loc. and see Vol. 1., Ch. 5; P.Mich.468.3-4 (=CPL 251), ante omnia opto te bene [u]alere, que m[ihi ma]xime uota [su]nt (for the Greek equivalent see P.Mich.476.3 ; CPL 304.2, opto deos ut bene ualeas que mea uota sunt. et before uotis in our text is emphatic and is picked up by et in line 8, i.e. it will be successful both because we want it and because you are worthy of success.
8. tu:over u there seems to be a short diagonal stroke like an apex mark. Similar marks are sometimes found in Latin papyri over long vowels (e. g. ChLA 418, speech of Claudius, CPL 247, private letter, late first century B.C.), cf. P.J. Parsons, JRS lxix (1979), p.134. We think it unlikely that the mark in line 3 can be an apex (see note) but we can at least explain a possible occurrence in the word tu. A single occurrence of the mark in a text might be thought unlikely, but it does occur once only in P.Hibeh 276.6 (=CPL 260).
9-11. consulari n(ostro) … occurres: the only governor named in the tablets is Neratius Marcellus (see No. 37.14-15 and note (225)); consularis appears without further identification in No. 30.5 (295) and perhaps in No. 36 (223). But given that these texts cannot be precisely placed within the period indicated by the archaeological context (i.e. the earlier part of the period A.D.95-105), it could easily be someone else.
n(ostro): it is certain that n (with superscript line) was preceded and followed by interpunct. This does not appear elsewhere in this text; we are, however, certain that the marks visible on the photograph are ink. This abbreviation also occurs in No. 30.5 (295), No. 34.3 (218) and perhaps also in No. 36 (223). It is found in papyri as early as A.D.103 (ChLA 215.25; when first editing this text in 1910 as P. Oxy.1022 Hunt remarked that it was unexpected, but it has turned up several times in papyri published subsequently cf. ChLA 270.6 note). Note that nostris in line 6 is not so abbreviated.
12. The second hand could be the same as that in No. 30 (295) but there is not really enough clear writing to make a proper comparison.
optamus: it would be easier to read op<t>amus, but such a mistake would be very unexpected; we have therefore decided to take the letter before amus as t rather than p and to suppose that p is represented by the slight traces at the right of o; m is ligatured to following u by raising the last stroke to the horizontal so that the letter looks more like n with a link stroke to u. Exactly the same thing has happened in the ligature mi in domine in line 13.
13. This line commences further out to the left, under the last letters of line 6. There is space between domine and bene which is more than enough to accommodate te.
ualere: a difficult reading. u is no more than a residual hook linked to a and only a small part of 1 and the final e is visible. In the context the only alternative which would seem possible is ualeas (for opto + subjunctive see lines 3-5), but we think the traces of the last letter suit e better than s and that the previous letter is more like r than a. We have considered the possibility that in lines 12 and 13 optamus frater / bene ualeas was written and that another line (which might have been Brocchus’s contribution after Niger wrote the first greeting) was then added, beginning domine and continuing beneath bene ualeas. However, ualere seems a better reading, and it seems very unlikely that a second greeting could have commenced with the word domine.
14. esse: this is probable but not certain and we think nore before it inescapable. All this suggests to us is in honore esse but the ink has faded so badly that we cannot read it.
Back. Ceri]ali: the traces are meagre and the reading little better than a guess. There would be room for praef(ecto) either before coh or on a separate line between this and the name. As is normal, the address in written on the back of col.ii.