chaser. Amulets intended to operate against a special ill would be in demand according to the prevalence of fear that the ill might come the way of the buyer; and those that mention the name of the wearer were, of course, inscribed on special order, just as a piece of jewelry may now be marked with the name or initials of the person for whom it is intended. Not only the inscription but the design also might be a special commission if it represented a deity particularly worshiped by the person bespeaking it, or if it was desired to conform to directions laid down by a master of magic, like the prescriptions for magical stones found here and there in the papyri. But the vast majority of amulets — though, as previously noted, the least interesting — were meant to be as good for one wearer as another, and it goes without saying that this kind was the most profitable to the dealer. Thus it comes about that several common types exist in hundreds of specimens, and that many inscriptions are known in several identical or only slightly differing examples. Wholesale production of cheap amulets is indicated by an interesting object in the collection of Mr. Seyrig (D. 328). It is a piece of steatite which formed one side of a mould intended for casting small amulets of lead, two at a time. I cannot date the lettering accurately, but it is certainly Byzantine, perhaps of the fifth or sixth century. The side preserved was inscribed with the first words of the ninety-first Psalm, which was in very common use as an apotropaic text, appearing on scores of Syrian bronze pendants and medals. A lead pendant produced from a smaller and much simpler mould is in the University of Michigan collection; it bears only the names Ρεφαηλ and Σαβαω.25
An inscription read by Zoega on a rock crystal of the Borgia collection presents a curious problem.26
There is no design, and the inscription, which is carved on both faces of the stone, consists mainly of the seven vowels variously combined, a few “characters,” and a few syllables that make no sense; but the inscription on the obverse ends thus: ΦΥΛΑΖΟΝ|ΤΟΝ ΑΕΙΝΑ ΤΟΝ ΦΟΡΟΥΝΤ|ΤΟ ΦΥΛΑΚΤΗΡΙΟΝ|ΤΤ ΤΟ ΑΦΑΡΤΗ. This should probably be read φύλαξον τὸν δεῖνα τὸν φοροῦντ‹α› τὸ φυλακτήριον τ‹οῦ›τ‹ο› τὸ ἀπάρτι, though it might be hard to parallel the abbreviation for τοῦτο. For the τὸ before ἀπάρτι we may compare τὸ νῦν. ἀφάρτη is a credible misspelling of ἀπάρτι; that word occurs in the New Testament meaning “henceforth” or “just now.” It seems in this place to be a substitute for ἤδη ἤδη ταχύ ταχύ, so common in papyrus spells. Now φύλαξον τὸν δεῖνα might stand in the gem cutter's copybook, and he would be expected to substitute an actual name if the customer desired it; φύλαξον τὸν φοροῦντα κτλ. would be the proper copy if the stone were to be offered for sale to any “bearer.” It is odd to find τὸν δεῖνα actually cut along with the other formula. Either the engraver wrongly combined the two or the wearer was supposed to say the charm to himself, substituting his own name for τὸν δεῖνα.27
Since many of these charms are like short prayers, it may be that the owner was expected to say them over at his pleasure.