1         

commonitorium: for this technical, quasi-legal term, found here for the first time in a curse tablet text, see TLL s v. It does not occur before the fourth century, when it becomes quite common, sometimes in the special sense of a petition to a superior (eg Symmachus, epistulae, i 68), which would seem to be the sense here.

2         

Mercurio: written over Marti Silvano (but deo Silvano has been retained in line 14), suggesting an alternative Roman ‘identification’ of the god of Uley; he is ‘identified’ as Mars in Uley-3, line 1 with note. The Celtic god Cocidius was variously ‘identified’ with Mars (RIB 602, 993, 2015, 2024) and with Silvanus (RIB 1578).

3         

muliere: why the sex (or status?) of Saturnina is specified is not clear. Sulis received texts from women identified only by their names (Tab Sulis 59, 60, 61, 97), and it does not follow that a (male) scribe is writing for Saturnina.

5         

circumvenit: unique in curse tablet texts as a synonym for the usual involavit; for the (perfect indicative) tense, cf Uley-1, line 13, with note.

5ff         

non ante laxetur (etc): a unique variant of the common formula (see Tomlin 1988, 65-6) that the thief is to be forced by ill health to return the stolen property. ante is redundant, unless quando[o] is intended for quam, when nissi would be redundant.

6         

nissi: for the divergent spelling, cf Uley-1 line 10, with note.

7         

s(upra)dictas: for the abbreviation cf deo s(upra)s(crip)to, from Ratcliffe-on-Soar (JRS 53 (1963) 123); and s(...) s (upra)s(crip)ti from Eccles (Britannia 17 (1986) 430, fig 1). It seems to be ‘clerical’ language, like infrascriptis in Tab Sulis 8 (and see Tomlin 1988, 64). ad fanum: for the formula that stolen property should be returned to the temple, cf Uley-4, line 10 with note) and ibid, 68.

8-9         

cf lines 16-17 below. These formulas are common in British curse tablets (see ibid, 67).

10         

deo s(upra)dicto: see note to line 7.

10-11         

tertiam partem [d]onat: cf line 15, tertia pars donatur: the ‘gift’ of a proportion of the stolen property to the god to secure his intervention is also found in Uley-4 and Uley-84. The formula is not found at Bath, but also occurs in Lydney-1 (one-half), PagansHill-1 (one-half), and J Roman Stud 53 (1963) 123 (Ratcliffe-on-Soar) (one-tenth).

12         

exsigat: the formula, the exaction of a debt by the god (who has been made part-creditor), is also found at Bath, Pagans Hill, and Ratcliffe-on-Soar (see Tomlin 1988, 64). The Vulgar spelling is also found in Tab Sulis 34.

13         

s(upra)s(crip)ta: see note to line 7.

14         

ACA quae per[didi]t: it is not clear from the electrotype whether there was another letter between ac and a and between per and t but it seems there was not. ACA quae might be understood as ac a quibus, but the conjunction ac (instead of et) is otherwise unknown in British curse tablet texts, and the grammar of a quaetertia pars is very harsh. The sense must be rerumtertia pars. per(didi)t for PERT makes sense (for the formula see Tomlin 1988, 65), and the verb seems to have attracted copying errors, cf Uley-4, line 3 (with note). perit is found in Britannia 17 (1986) 431, no 2 (Eccles), but, unless it is corrupt, in the sense of the thief ‘perishing’.

deo Silvano: see note to line 2.

15         

tertia pars: see note to lines 10-11; presumably the same one-third, the god’s name ‘Silvanus’ being retained in error.

16         

exsigat: see note to line 12.

16-17         

cf lines 8-9 above (with note). The last word is almost lost; evidently it was a verb in the present tense, third person singular, but there is no obvious candidate.