salutem: this is written at the right-hand end of the line and we assume that a preceding lost line contained the beginning of the letter, Claudia Seuera Lepidinae suae, written across the whole width of the double leaf (Claudia is abbreviated in 291.i.1, written by a different hand).
ego: go is certain; before it we should probably simply read e and take the other marks as offset. For another letter beginning with ego see 265.3. We would not rule out as an alternative reading ergo; its position at the beginning of a sentence is quite normal, cf. P.Mich.VIII 469.17 (= CEL 144); at the beginning of a letter it suggests response to, or resumption of, something discussed in an earlier letter (OLD, s.v. 5).
locuta fueram: see Adams (1994). There is a mark over locuta which is not ink.
at te: also in ii.3, cf. P.Mich.VIII 472.4, 17 (= CEL 147), Adams (1977), 25-9.
mihi <i>ta corde: we have been unable to find a way of resolving the difficulties in this line without supposing a scribal error due to haplography. The reading of mihita is virtually certain (there is no serious doubt about a). Following that cor is clear and, after a doubtful letter, esem is equally clear; as the broken letter thereafter suits p very well, we think semp[er is certain. The doubtful letter after cor could be t, but we cannot suggest what corte would mean (unless it were to be understood as being for corde; the scribe does write at te in lines a.i.3 and b.ii.3 but this use of t for d is much less surprising). We think corde defensible as a reading; for d written with a very small bow cf quaedam in b.ii.4. Since tacorde makes no sense and ita fits very well, we can only suppose that this was what was intended and that the second i has been omitted. For corde perhaps see Virgil, Aen. 6.675, si fert ita corde uoluntas, Plautus, Capt. 420, uideas corde amare inter se, and cf. TLL IV 940 which seems to suggest that it could mean ex animo.
li]citum uná: only the top of the letter which we have read as c is visible but it compares well with the example in actura (b.iii.4) and the word fits the context very well. We imagine that the following line (b.ii.1) began with cum plus name or description or both (e.g. cum Candido seruo, or perhaps joining a military party, cum militibus/centurionibus cohortis). An alternative way of reconstructing the sense of what follows is to suppose that li]citum is the end of a sentence and that in the subsequent sentence Severa is saying "therefore I shall try to visit you, in company with X, by whatever means I can". It seems less likely that uná is nominative (to which the apex is no objection cf. necessariá, b.ii.4), meaning "alone". In what follows she could be saying either that there is some essential business which she needs to discuss with Lepidina face-to-face rather than by correspondence or that there is some essential business which prevents Brocchus from accompanying her.
at te: see a.i.3 note.
There is a mark on the right-hand half of the leaf, after the gap, which cannot be the top of e in this hand; it can hardly be the top of c or s either (which in any case make no sense here). Unless it is just an offset, which looks improbable, it is an apex. For an apex over a diphthong see CEL 72.7, 81.11. Alternatively, there might perhaps be enough space to accommodate qua[e á].
Perhaps a gerundive (e.g. agenda sunt) to end the sentence, then a new sentence beginning, for example, per frat]/rem or per familia]/rem.
meas: this seems to us the likeliest reading although only a is clear; neither a me nor duas can be read.
sim actura: for the future-exponent see Adams (1977), 49. There might be an apex over the a at the end of actura, though this may be just offset.
Col.iv will have been written on the left-hand half of this diptych and is wholly lost. We must therefore reckon with the loss at this point of 4-5 lines in col.iv before the first line of col.v (of which only exiguous traces remain).
.ra: there is a mark, like an apex, over the r but we think this is probably not ink. In view of mansura following we expect another future participle here; although we could just possibly read ura the word division is unlikely.
For Briga cf. 190.c.38-9 and note, domini Brigae man[se-]/runt; it is obviously the name of a place in the vicinity of Vindolanda. It remains uncertain whether Briga was the home base of Brocchus and Severa; the fact that Severa explicitly states her intention to remain there might suggest that Lepidina would not necessarily expect her to be there. It should also be noted that mansura might mean "intending to stop off at" or "intending to lodge at" (OLD, s.v.2a) and, if this were the case, it would imply that Briga was not her home base.
Cf. 291.ii.9, where a me is omitted.
It is very unclear on the photograph which marks are ink and which are not and this makes the readings particularly difficult.
The first visible letter is certainly e and the spacing permits [ual]e (cf. 291.ii.12, 293.1). The next letter is almost certainly m and the line clearly ends ror; we therefore expect mea soror, but it is very hard to read this.
karissima: the reading of the first two letters is most uncertain.
et anima / ma: we expect et anima mea or et mea anima, but find it impossible to read either. We give what we think is the least problematical reading but would stress that it is very hard to read the dotted letters. For the possessive adjective in this form see Adams (1994). For the use of anima in the closure see 291.ii.12 with the note, cf. 293.1 note. For desideratissima see CIL 6.21974.6-8, coniugi carissim[ae] animae desideran[tissi]mae (cf. TLL VI.1 710).
The traces might belong to [ua]le.
As in 291.back 16, coniugi is omitted. It is noteworthy that the last three letters of Lepidinae are written in ordinary cursive, whilst the first six letters in the previous line are in the spindly script associated with addresses. There may be traces further to the right; if so, presumably from praef(ecto) coh(ortis).
a Seuera: only the tops of the last two letters survive. Thereafter there is a trace of the top of an ascender which is compatible with b and it is difficult to imagine that we could have anything other than B[rocchi.