the /Musa Stratonis/ has several which happily in no way belong to it.
There is no doubt a certain charm to the very confusion of the order,
which gives great variety and unexpectedness; but for practical
purposes a more accurate classification is desirable.

The Anthology of Planudes attempts, in a somewhat crude form, to
supply this. Each of the six books, with the exception of the
{Erotika}, which remain as is in the Palatine Anthology, is subdivided
into chapters according to subject, the chapters being arranged
alphabetically by headings. Thus the list of chapters in Book I.
begins, {eis agonas}, {eis ampelon}, {eis anathemata}, {eis
anaperous}, and ends {eis phronesin}, {eis phrontidas}, {eis khronon},
{eis oras}.

On the other hand, Brunck, in his /Analecta/, the arrangement of which
is followed by Jacobs in the earlier of his two great works, recast
the whole scheme, placing all epigrams by the same author together,
with those of unknown authorship at the end. This method presents
definite advantages when the matter in hand is a complete collection
of the works of the epigrammatists. With these smaller, as with the
more important works of literature, it is still true that a poet is
his own best commentator, and that by a complete single view of all
his pieces we are able to understand each one of them better. A
counter-argument is the large mass of {adespota} thus left in a heap
at the end. In Jacobs there are upwards of 750 of these, most of them
not assignable to any certain date; and they have to be arranged
roughly by subject. Another is the fact that a difficulty still
remains as to the arrangement of the authors. Of many of the minor
epigrammatists we know absolutely nothing from external sources; and
it is often impossible to determine from internal evidence the period,