unknown date; the most interesting being Alpheus of Mitylene, who from the style and contents of his epigrams seems to have lived about the time of Hadrian, but may possibly be an Augustan poet. Strato is mentioned by Diogenes Laërtius,[18] who wrote at the beginning of the third century; and his own epigram on the physician Artemidorus Capito,[19] who was a contemporary of Hadrian, fixes his approximate date.

How far we possess Strato's collection in its original form it is impossible to decide. Jacobs says he cannot attempt to determine whether Cephalas took it in a lump or made a selection from it, or whether he kept the order of the epigrams. As they stand they have no ascertainable principle of arrangement, alphabetical or of author or of subject. The collection consists of two hundred and fifty-nine epigrams, of which ninety-four are by Strato himself and sixty by Meleager. It has either been carelessly formed, or suffered from interpolation afterwards. Some of the epigrams are foreign to the subject of the collection. Six are on women;[20] and four of these are on women whose names end in the diminutive form, Phanion, Callistion, etc., which suggests the inference that they were inserted at a late date and by an ignorant transcriber who confused these with masculine forms. For all the epigrams of Strato's collection the Anthology is the only source.

In the three hundred years between Strato and Agathias no new Anthology is known to have been made.

The celebrated Byzantine poet and historian Agathias, son of Mamnonius of Myrina, came to Constantinople as a young man to study law in the year 554. In the preface to his History he tells us that he formed a new collection of recent and contemporary epigrams previously unpublished,[21] in seven books, entitled {Kuklos}. His proem to the Cyclus is extant.[22] It consists of forty-six iambics followed by eighty-seven hexameters, and describes the collection under the symbolism no longer of a flower-garden, but of a feast to which different persons bring contributions ({ou stepsanos alla sunagoge}), a metaphor which is followed out with unrelenting tediousness. The piece is not worth transcription here. He says he includes his own epigrams. After a panegyric on the greatness of the empire of Justinian, and the foreign and domestic peace of his reign, he ends by describing the contents of the collection. Book I. contains dedications in the ancient manner, {os proterois makaressin aneimena}: for Agathias was himself a Christian, and indeed the old religion had completely died out even before Justinian closed the schools of Athens. Book II. contains epigrams on statues, pictures, and other works of art; Book III., sepulchral epigrams; Book IV., epigrams "on the manifold paths of life, and the unstable scales of fortune," corresponding to the section of {Protreptika} in the Palatine Anthology; Book V., irrisory epigrams; Book VI. amatory epigrams; and Book VII., convivial epigrams. Agathias, so far as we know, was the first who made this sort of arrangement under subjects, which, with modifications, has generally been followed afterwards. His Anthology is lost; and probably perished soon after that of Cephalas was made.

Constantinus Cephalas, a grammarian unknown except from the Palatine MS., began again from the beginning. The scholiast to the Garland of Meleager in that MS., after saying that Meleager's Anthology was arranged in alphabetical order, goes on as follows:--"but Constantinus, called Cephalas, broke it up, and distributed it under different heads, viz., the love-poems separately, and the dedications and epitaphs, and epideictic pieces, as they are now arranged below in this book."[23] We must assume that with this rearranged Anthology he incorporated those of Philippus and Agathias, unless, which is not probable, we suppose that the Palatine Anthology is one enlarged from that of Cephalas by some one else completely unknown.

As to the date of Cephalas there is no certain indication. Suidas apparently quotes from his Anthology; but even were we certain that these quotations are not made from original sources, his lexicon contains entries made at different times over a space of several centuries. A scholium to one of the epigrams[24] of Alcaeus of Messene speaks of a discussion on it by Cephalas which took place in the School of the New Church at Constantinople. This New Church was built by the Emperor Basil I. (reigned 867-876). Probably Cephalas lived in the reign of Constantine VII. Porphyrogenitus (911-959), who had a passion for art and literature, and is known to have ordered the compilation of books of excerpts. Gibbon gives an account of the revival of learning which took place under his influence, and of the relations of his Court with that of the Western Empire of Otto the Great.

The arrangement in the Anthology of Cephalas is founded on that of Agathias. But alongside of the arrangement under subjects we frequently find strings of epigrams by the same author with no particular connection in subject, which are obviously transcribed directly from a collected edition of his poems.

Maximus Planudes, theologian, grammarian, and rhetorician, lived in the early part of the fourteenth century; in 1327 he was appointed ambassador to the Venetian Republic by Andronicus II. Among his works were translations into Greek of Augustine's City of God and Caesar's Gallic War. The restored Greek Empire of the Palaeologi was then fast dropping to pieces. The Genoese colony of Pera usurped the trade of Constantinople and acted as an independent state; and it brings us very near the modern world to remember that while Planudes was the contemporary of Petrarch and Doria, Andronicus III., the grandson and successor of Andronicus II., was married, as a suitable match, to Agnes of Brunswick, and again after her death to Anne of Savoy.

Planudes made a new Anthology in seven books, founded on that of Cephalas, but with many alterations and omissions. Each book is divided into chapters which are arranged alphabetically by subject, with the exception of the seventh book, consisting of amatory epigrams, which is not subdivided. In a prefatory note to this book he says he has omitted all indecent or unseemly epigrams, {polla en to antigrapso onta}. This {antograpso} was the Anthology of Cephalas. The contents of the different books are as follows:

Book I.--{Epideiktika}, in ninety-one chapters; from the {Epideiktika} of Cephalas, with additions from his {Anathematika} and {Protreptika}, and twelve new epigrams on statues.

Book II.--{Skoptika}, in fifty-three chapters; from the {Sumpotika kai Skoptika} and the {Mousa Stratonos} of Cephalas, with six new epigrams.

Book III.--{Epitumbia}, in thirty-two chapters; from the {Epitumbia} of Cephalas, which are often transcribed in the original order, with thirteen new epigrams.

Book IV.--Epigrams on monuments, statues, animals, and places, in thirty-three chapters; some from the {Epideiktika} of Cephalas, but for the greater part new.

Book V.--Christodorus' description of the statues in the gymnasium called Zeuxippus, and a collection of epigrams in the Hippodrome at Constantinople; from appendices to the Anthology of Cephalas.

Book VI.--{Anathematika}, in twenty-seven chapters; from the {Anathematika} of Cephalas, with four new epigrams.

Book VII.--{Erotika}; from the {Erotika} of Cephalas, with twenty-six new epigrams.

Obviously then the Anthology of Planudes was almost wholly taken from that of Cephalas, with the exception of epigrams on works of art, which are conspicuously absent from the earlier collection as we possess it. As to these there is only one conclusion. It is impossible to account for Cephalas having deliberately omitted this class of epigrams; it is impossible to account for their re-appearance in Planudes, except on the supposition that we have lost a section of the earlier Anthology which included them. The Planudean Anthology contains in all three hundred and ninety-seven epigrams, which are not in the Palatine MS. of Cephalas. It is in these that its principal value lies. The vitiated taste of the period selected later and worse in preference to earlier and better epigrams; the compilation was made carelessly and, it would seem, hurriedly, the earlier part of the sections of Cephalas being largely transcribed and the latter part much less fully, as though the editor had been pressed for time or lost interest in the work as he went on. Not only so, but he mutilated the text freely, and made sweeping conjectural restorations where it was imperfect. The discrepancies too in the authorship assigned to epigrams are so frequent and so striking that they can only be explained by great carelessness in transcription; especially as internal evidence where it can be applied almost uniformly supports the headings of the Palatine Anthology.

Such as it was, however, the Anthology of Planudes displaced that of Cephalas almost at once, and remained the only MS. source of the anthology until the seventeenth century. The other entirely disappeared, unless a copy of it was the manuscript belonging to Angelo Colloti, seen and mentioned by the Roman scholar and antiquarian Fulvio Orsini (b. 1529, d. 1600) about the middle of the sixteenth century, and then again lost to view. The Planudean Anthology was first printed at Florence in 1484 by the Greek scholar, Janus Lascaris, from a good MS. It continued to be reprinted from time to time, the last edition being the five sumptuous quarto volumes issued from the press of Wild and Altheer at Utrecht, 1795-1822.

In the winter of 1606-7, Salmasius, then a boy of eighteen but already an accomplished scholar, discovered a manuscript of the Anthology of Cephalas in the library of the Counts Palatine at Heidelberg. He copied from it the epigrams hitherto unknown, and these began to be circulated in manuscript under the name of the Anthologia Inedita. The intention he repeatedly expressed of editing the whole work was never carried into effect. In 1623, on the capture of Heidelberg by the Archduke Maximilian of Bavaria in the Thirty Years' War, this with many other MSS. and books was sent by him to Rome as a present to Pope Gregory XV., and was placed in the Vatican Library. It remained there till it was taken to Paris by order of the French Directory in 1797, and was restored to the Palatine Library after the end of the war.

The description of this celebrated manuscript, the Codex Palatinus or Vaticanus, as it has been named from the different places of its abode, is as follows: it is a long quarto, on parchment, of 710 pages, together with a page of contents and three other pages glued on at the beginning. There are three hands in it. The table of contents and pages 1-452 and 645-704 in the body of the MS. are in a hand of the eleventh century; the middle of the MS., pages 453-644, is in a later hand; and a third, later than both, has written the last six pages and the three odd pages at the beginning, has added a few epigrams in blank spaces, and has made corrections throughout the MS.

The index, which is of great importance towards the history not only of the MS. but of the Anthology generally, runs as follows:--

{Tade enestin en tede te biblo ton epigrammaton

A. Nonnou poirtou Panopolitou ekphrasis tou kata Ioannen agiou euaggeliou.