The practice of inscribing works of art with the names of persons and objects represented was one of some antiquity in Greece. The earliest instance of which we have historical record is the chest of Kypselos, which dated from the beginning of the sixth century B.C., and concerning which Pausanias[2058] tells us that “the majority of the figures on the chest have inscriptions written in the archaic characters; and some of them read straight, but other letters have the appearance called by the Greeks ‘backwards-and-forwards’ (βουστροφηδόν), which is like this: at the end of the verse the second line turns round again like a runner half through his course. And any way the inscriptions on the chest are written in a tortuous and hardly decipherable fashion.” There is, however, no mention of inscribed vases until a much later date; Athenaeus speaks of a cup with the name of Zeus Soter upon it, also of γραμματικὰ ἐκπώματα, or cups with letters on them.[2059] Inscriptions on Greek vases are found in comparatively early times, even prior to the date to which the chest of Kypselos is attributed. This question will receive more attention subsequently; meanwhile, we may point out some of the ways in which they have proved important in the study of archaeology. In the first place, they were originally among the principal, perhaps the strongest, arguments in the hands of Winckelmann, Sir W. Hamilton, and the other upholders of the true origin of Greek vases against Gori and the other “Etruscans” (see Vol. I. p. 19). They are, in fact, if such were required, an incontestable proof of Greek manufacture. Secondly, in more modern times, they have been of inestimable value in enabling scholars to classify the early vases according to their different fabrics. The alphabets of the different cities and states being established by inscriptions obtained from trustworthy sources or found _in situ_, it was an easy matter to apply this knowledge to the vases. In Chapters VII.-VIII. numerous instances have been given of the value of this evidence (see also below, p. 247 ff.), perhaps the best being that of the Chalcidian class, for which the inscriptions have been a more important criterion even than style. Thirdly, the inscriptions are sometimes of considerable philological value. Those on Attic vases may fairly be said to represent the vernacular of the day; and thus we learn that the Greeks of the Peisistratid age spoke of Ὀλυττεύς, not Ὀδυσσεύς, and of Θῆσυς, not Θησεύς; that they used such forms as υἱύς for υἱός,[2060] and πίει for πίε (see below, p. 255). Traces of foreign influence in the inscriptions, as in the frequently occurring Doric forms, imply that many of the vase-painters were foreigners, probably of the metic class. We shall also see that one class of inscriptions gives some interesting information on the subject of the names and prices of vases in antiquity.