These difficulties are not insuperable, but first I wish to refer to another interpretation, which has enjoyed great popularity. There is no doubt that the Greeks were acquainted, and at an early date, with the alphabetical arrangement of titles. The Oxyrhynchus arguments to Menander’s plays, for example, seem to have been arranged in accordance with this principle. The objection that there would be no point in recording numerals derived from an alphabetical system for the reason that it would be as easy to turn to a given play by means of its initial letters as by means of its number is invalid because in alphabetical lists the Greeks ignored all letters except the first. For example, fifteen of Euripides’ extant titles begin with alpha, and there was no a priori method of knowing which of the fifteen places available the _Alcestis_ would occupy (Fig. 80).[380] It becomes necessary, then, to examine the alphabetical explanation without prejudice, and fortunately it is now possible to reach an incontrovertible conclusion. The numerals have never lent themselves cordially to this interpretation, but the final _coup de grâce_ was delivered by the recent discovery of the numeral for Menander’s _Imbrians_. Menander is said to have written from one hundred and five to one hundred and nine pieces, but only eighty-six titles are now known. Fifty-one of these, however, have initial letters which come after iota in the Greek alphabet. Now the smallest restoration which is possible for the Menander numeral is seventy-one, and seventy-one plus fifty-one make one hundred and twenty-two, or thirteen more than the largest number recorded by any authority as the aggregate of Menander’s works. Therefore the alphabetical explanation must be rejected. [Illustration: FIG. 80.—The Villa Albani Statue of Euripides in the Louvre with the Beginning of an Alphabetical List of His Plays. See p. 332, n. 1]