We are reasonably confident of the reading, although p and r of primum are abraded and difficult (and the final m is probably crossed by a descender from the line above).
].arem: the trace after the break can be read as i and we suggest the restoration [eum] f[amil]iarem. At the end of the line, where the writing is very blurred, we could read sub as an alternative, but do not see how to construe sub-/misseras here. We can read the vertical as the left-hand stroke of h and mihi is to be expected if misseras in the following line is correct.
The first two letters of misseras are difficult to read, especially m. After non we could read quare which would justify the subjunctive sciat and might suggest something like "I see no reason why he should know ...". Such a broken context requires great caution, however.
For debts between soldiers see Tomlin (1992), 148, note 33 and perhaps cf. 193, 206.
We could supply per at the beginning of the line and suppose that the debt was to be returned through Candidus and Natalis, but this would certainly not account for all the loss. We might obtain a better fit by supposing that a new sentence began here, e.g. [(re)mittam Can]didum et Natalem, and perhaps a third name at the beginning of the next line. For persons in the Vindolanda text with the common name Candidus see 301.i.1 note; the name is far too common to allow any identification and this Candidus may well come from elsewhere. The name Natalis does not occur elsewhere in the tablets.
Vir..cium: the most plausible reading is Viriocium; this is reported by A.R.Birley (1991), 89 as Viriocus; the closest parallel we have found for this is CIL 13.4514, Viriaicus (cf. Birley, loc.cit., Weisgerber (1969), 232).
excipias bene: cf. Petronius, Sat. 100.4, quam bene exulem exciperem.
corist..st: A.R.Birley (1990a), 18 (cf. VRR II, 37) reads "Corieptusti(um) (?)". The reading is difficult since the traces after the first six letters are very abraded and it is uncertain how much more writing there is. We regard the reading of Coris followed by t as quite certain; coriep- results from taking the left-hand part of the top of t as the cross bar of an e. Since the word or words are placed in the top-left corner of the leaf it is natural to take them as a place-name in an address (see above, pp.43-5). Coris with something following suggests a name of the pattern Coria plus a tribal name, of which several are attested (see PNRB 317-20, arguing that all the coria-names in Britain and on the continent derive from the Celtic word for "hosting-place, tribal centre" and that some eventually became assimilated to Latin curia). If this is correct, it is very tempting to connect the name on the Vindolanda letter with Curia Textoverdorum, attested by RIB I 1695 and certainly in the region of Vindolanda, perhaps its vicus (see PNRB 470-2). The form of the name is problematical (see PNRB 471), but it is possible to suggest that we here have Coris Texsto followed by indeterminate traces.
Cessaucio: the reading is certain, but the gentilicium is unattested. We have found: Cessicius (NPEL, CIL 3.7448 (Moesia)); Cassicius (CIL 6.2675); Cessitius (LE 428); Cacussius, Weisgerber (1969), 141. A.R.Birley (1991), 92 (cf. VRR II, 37) cites the cognomen as Morin[o (attested once in Gallia Belgica, see NPEL), which might evoke a connection with the Morini of north-west Gaul, but this reading is very difficult. The second visible vertical has no slant and suggests that we should read Ni rather than M; the two letters before the break do seem to be in; perhaps Nigrin[o, although there hardly seems to be room for g.
dup(licario): this does not occur elswhere in the Vindolanda texts; cf. CEL 83 = ChLA X 424. For auxiliary duplicarii see Breeze (1974), 278-86.
a Tullioni: the form of a is interesting because a second descender is written more horizontally than is normal and almost makes a cross-bar. For the ablative in -i see Tab.Vindol.I, p.74. A.R.Birley (1990a), 18 (cf. VRR II, 37) has cited the name as Pullio but this is not correct; the cross-stroke at the top extends on both sides of the vertical. The name Tullio occurs twice in 184.ii.29, iii.31 presumably borne by people at Vindolanda. The name is reasonably well-attested but occurs only twice in Gallia Belgica according to NPEL.