PREFACE

a pointed bracket indicates that the letters within
it are net or cannot have been in the MS. The
emendations and restorations for which [ believe
myself to be responsible are marked # in the notes.
All restorations have been checked where possible
by a palacographical method explained and exempli-
fied in various articles in the periodicals. Briefly, it
consists of the tracing of letler-groups from photo-
graphs of the extant portions of the papyrus or
véllum Ms. Suggestions are rejected which, when
traced out by this method, are shown, with all
reasonable allowance made for variation in the size
of the letters and the spaces between them, to
exceed or fall short of the requirements of the gap:
and where the gap is bounded on the left by an
imaginary marginal line, all suggestions are made to
correspond—again with all reasonable allowance
made—in what I may call for convenience ‘ written
leneth, Scholars who have not tried this method
will be surprised, when they do, at the way in which
it reduces the possibilities. One instance must
suffice. In the first lines of Aleaeus 27, the letters
A of παίδων, Al of πρώϊτᾳ. TO betore yap, and IC of
GAX Jans, come immediately under one another. Metre
requires Geo supplementary syllables in lines | and
2. three in line 5, and one in line 4. ΑἸ] these
supplements must correspond in written length not
only with one another but with any suggestions
made for the four subsequent lines, and w bos they
are made the result must be not only a passage
satisfactory in metre, grammar, dialect and sentiment,
but something which Aleaeus might have written.
1 should add here that so far as I have found it
practicable my work on the new Fragments is based

is