The sherd is a small, nearly flat piece. It looks like a fragment of a bigger sherd: ends of hastae can be seen in the breaks. The sherd is crossed by an inscribed straight line slightly below the mid-height of the sherd. The Inventory card in the Hermitage notes that Krasotkina and Bilimovich under the direction of I.I. Tolstoy, processed this material for the museum records in March 1949. Their proposed reading was ἡ γλή[νη] | [- -]ν δι-| οίχο and their translation: "Pupil (of an eye).... perish!" where διοίχο is understood as διοίχου. This restoration makes little sense, and I think it is influenced by I.I. Tostoy's tendency (Греческие граффити древних городов Северного Черноморья. Leningrad, 1953) to see dedicatory insriptions and harmful magical inscriptions in fragmentary graffiti. More importantly, it does not correspond to the preserved letters.

A horizontal divider line cuts the sherd into two uneven halves. On each side of the divider, there are lines of text. Line 1. The first letter must be a tau: the vertical and right half of the horizontal are clear. The second letter is eta. The third must be a nu, with raised right hasta. The following is either a lambda or alpha, followed by eta with long verticals. There are markings below this line of text, but it is impossible to make sense of them; it is not clear if they form part of the text or not; to my eye, they look like slips of the writing implement as they seem to be lined up below the strokes of the line above. Although the proposed seems to be the safest interpretation of letter strokes, there is a slight possibility that what I read as a lambda was actually an alpha (the letter strokes have chipped the surface so much that the damage might have covered a short horizontal of an alpha. It also cannot be excluded that the short stroke that I restore as a possible tau belonged to a horizontal of a cross-wheel theta, which extended beyond the circle. Allowing these alternatives to tau and lambda, we would end up with [Ἀ]θ̣ηναη[- -], which is certainly tempting, but should probably be resisted for that very reason.

The difficulties mount for the rest of the graffito. The direction of sigma suggests that the sherd may have been flipped upside down, and the first line in the second half would be the one on the edge of the fragment, however, it is also possible that the sherd was not flipped, but the direction of writing switched to retrograde; in that case, the line next to the divider would be Line 2. Another implication of the latter arrangement would be that what looks like a dzeta would have to be read as a tau.

Line 2. I prefer the option of a flipped sherd, which might be suggested by the existence of the horizontal divider line. The first two letters ΟΣ seem certain to me, and might suggest an ending of a word or name. The following vertical has a short horizontal at the bottom suggesting a zeta, followed by a squarish omicron, of the same shape as before the sigma. There appear to be remains of a letter to the right of omicron.

A long diagonal that stretches from the straight divider line into the space between omicron and sigma is hard to see as a letter stroke. Line 3. To the right of that diagonal, after some space - two letters are written closely together: iota and either a delta or omicron. I opt for the latter because there seems to be a rounded stroke that merges with the straight divider line. After some space, another letter in that line appears to be a nu (the extension of the diagonal to the right beyond the right vertical of nu looks to me like a slip of the writing implement.

What category of text we are dealing with is impossible to say with any degree of certainty. The use of a sherd, small letter size, four lines of text, the use of a noun in the accusative, a possible substantive ending in -ΟΣ suggest a message or note, but we cannot exclude some sort of an account. Either interpretation would be supported by the residential context of find.