University of Virginia Library


Page 77

Milton and the Harvard Pindar
Maurice Kelley and Samuel D. Atkins

In the Houghton Library at Harvard is a copy of Pindar (Sumner 123) commonly supposed to have belonged to John Milton. Notice of this volume first appears in the Sotheby sale catalogue of the library of J. B. Inglis (1871), where it is described as follows:[1]

1588 MILTON (JOHN). Pindari Olympia, Pythia, Nemea, Isthmia, Graecè, Jo. Benedictus innumeris mendis repurgavit, metaphrasi recognita, latina paraphrasi addita, half russia 4to. Salmurii, 1620.

This most precious and inestimable volume formerly belonged to the divine Milton. From a note on the title page we learn that he purchased the volume Novemb. 15, 1629, pret. 9s., and at the end, the dates of the period he was occupied in reading it, viz. from June 17 to Sept. 28, 1630. The margins throughout bear his Notes, many of them being very copious, and at the end he has added an Alphabetical Index, occupying two closely written pages, of all the Authors cited (except Homer and Pindar), with references to the different places where they are mentioned in the Annotations, a work of immense perseverance, and which no one except with the greatest labour of love would have done.

We cannot speak too highly of this book, nor can we conceive any thing to be more esteemed or revered than this copy of the works of the Prince of Lyric Poets annotated throughout by the Author of the Immortal "Paradise Lost," in his own handwriting.

The sale of this book and the purchaser's authentication of it are recounted by B. M. Pickering in an autograph note now pasted in before the front flyleaf of the volume:

This book (Pindar) I purchased at Messrs. Sotheby's Aug 5th 1871 my opponent being Mr. Addington the well known collector of autographs — Mr. Waller the dealer in autographs told me after the sale that he had no


Page 78
doubt about the notes being Milton's writing. After its purchase I had an opportunity afforded me of comparing this volume with the Lycophron bought for Mr. Forster at Lord Charlemont's sale The writing compared very satisfactorily therewith when the nature of the paper on which the Lycophron is printed is taken into consideration it being nearly like blotting paper & very difficult to write upon & in another way this Lycophron afforded a strong piece of circumstantial evidence as in this Pindar certain passages of Lycophron occurring at certain pages are referred to and all these passages are to be found exactly at the pages as referred to in the MS notes in this book The date given as the time occupied in the perusal (see end of book before Errata) viz Jun. 17, 1630 to Sept 28-1630 was part of a forced interval of leisure which Milton was compelled to take during his University career owing to the Masters fellows & students all having fled from Cambridge during the ravages of a fearful epidemic from April to end of Nov 1630. Cambridge was completely deserted (see Massons Life of Milton Vol 1 page 200) I gave at the sale £41 for the volume Mr. Addington the under-bidder having bid £40
B M Pickering

London 196 Piccadilly
Nov 11. 1872

Subsequent to this note — as a clipping from a catalogue now pasted on the front flyleaf indicates — the volume was again offered for sale as Milton's copy, and passed into the possession of Senator Charles Sumner. As a part of his bequest, it was received by Harvard April 28, 1874.

Since that time, students of Milton have unanimously accepted the claims of the sale catalogue and Pickering. Hanford,[2] for instance, lists the Harvard Pindar as one of the eight books that bear Milton's "unquestionably genuine autograph." The Columbia Milton[3] pronounces the notes "almost without exception in his own hand," and prints 122 of them with translations. And two other scholars have used the annotations to reconstruct Milton's Greek studies at Cambridge,[4] to identify editions of other classical authors that he read,[5] and to question the practice of dating Milton's handwriting on the basis of the form of e that it contains.[6] Such general acceptance and confident use would seem to establish beyond question Milton's ownership of the Harvard volume; but this ownership is precisely what we propose to


Page 79
query, and our reasons for doing so — set forth briefly under two chief heads of form and content — are as follows.

First, the Harvard Pindar does not contain Milton's signature; and lacking this, the volume can be safely accepted as Milton's only if the annotations significantly resemble the notes found in other books unquestionably owned and annotated by Milton. Six such books are extant: Milton's Dante-Della Casa-Varchi volume (New York Public Library), his Aratus (British Museum), his Euripides (Bodleian Library), his Lycophron (Mr. Adrian van Sinderen), his Dio Chrysostom (Chapter Library, Ely Cathedral), and his Heraclides (University of Illinois Library).[7] A comparison of these volumes reveals, however, that the annotations in the six Milton books consistently resemble each other, but differ in numerous ways from the notes in the Harvard Pindar.

A specific example of this similarity and disagreement appears in the seven records of purchase reproduced in Plate I. In the Pindar, the date and price appear on the same plane in opposite upper corners of the title page; in the six authentic volumes, signature, price, and date appear in varying order one below the other. In the Pindar, the abbreviation of "pretium" is "pret:"; all six Milton books read "pre:". In the Pindar, the price in even shillings is emphasized by a dash and a zero in the pence column; this practice Milton does not follow in the three instances where no pence are involved.

In the manuscript notes, as Plates II-III illustrate, further differences exist. As annotation marks, Milton uses either a grave accent, an asterisk, or a small x, and regularly places these marks on a plane above, if not directly over, the handwritten correction or note.[8] In the Pindar, the annotator often underlines the text, uses no grave accent, and on the occasions that he does use the asterisk or small x, he tends to place them on the same plane as his following note.[9] To call attention to a passage, Milton employs a large X or a perpendicular pen mark resembling a large parenthesis;[10] for the same purpose, the Pindar scribe uses a number of signs not found in the Milton volumes: #, double π, two horizontally parallel wavy lines, a pointing hand, three pyramidically arranged dots or a trefoil above a descending, wavy tail, and two trefoils joined by a sublinear loop.[11]


Page 80

Significant differences also appear in the Greek script. As Plates II-III show, the γ of the Pindar notes tilts so markedly to the right that on one occasion the Columbia editors[12] misread it as a 2; Milton's γ is more perpendicular and often suggests a large V dropped well below the line.[13] In executing μ, both the Pindar scribe and Milton begin by writing a Roman u; but in completing the letter, the Pindar scribe uses a descending wiggle, Milton a straight line generally terminating in a small hook to the left.[14] In the Pindar notes, the β is the familiar form resembling a tailed B; in Milton's annotations, the letter frequently appears in a cursive form slightly similar to an italicb.[15] Conversely, the Pindar notes often show a cursive θ that suggests a large 9 tilted to the right; in Milton's notes, the θ is the conventional upright oval and bar.[16]

Differences in form, therefore, indicate that the Pindar notes are not in Milton's autograph; and a second kind of difference — this time in content — argues with similar force against attributing the annotations to an amanuensis writing at Milton's dictation.

In his annotations of Greek poetry, Milton focusses his attention on the text, and his studies rarely pass beyond the translation, scholia, and commentary in the volume before him. Twelve of Milton's thirty-seven Aratus notes,[17] it is true, derive from collations of his Morel edition with the de Gabiano and Stephanus texts, and three others come "ex aliis editionibus", from Stephanus, Thesaurus Graecae Linguae, and perhaps from Grotius, Syntagma Arateorum; but in all these fifteen instances, Milton's interest is in text rather than in commentary — in finding the best or an equally good alternative reading; and from this purpose he varies only twice: once to quote Ovid's tribute to Aratus and once to give a parallel from Lucretius. Similarly, Milton's annotations of Lycophron contain two references to the Stephanus Thesaurus


Page 81
(pp. 21, 60), but here again his interest is textual, and his sixty-three remaining notes contain only three references to works outside his volume: to the scholia on the Argonautica (p. 27), to the Metamorphoses of Ovid (p. 145), and to the Ion of Euripides, (p. 156).[18] In over 500 annotations of Euripides, Milton's focus is even sharper: except for a single reference to "Scaliger in prooemio Manilium" (II, 69), his notes fail to indicate that he went beyond the covers of his Stephanus edition.[19]

In contrast, the Pindar annotator ranges freely beyond his volume. An instance of the "hors load of citations" that he deposits at his reader's door is a cluster of five notes entered on pages 2 and 4 concerning Olympia I, 6. The text (p. 2) reads ϕίλον ἦτορ, and the printed commentary (p. 4-5) has a six line note establishing that ϕίλος in this context is equivalent to ἵδιος and denotes "one's own." To this already adequate note, the Pindar annotator adds ten more citations or quotations from Homer, from Eustathius on Homer, and from a renaissance elegy that can have no primary authority in a question of Classical usage. Elsewhere, the Pindar annotator further supplements the printed commentary on Olympia I with quotations from or references to Lucian (p. 1), Aristotle (p. 2), Lycophron (pp. 3, 11, 35), Anacreon (p. 7), Horace (pp. 7, 10, 12, 22, 30), Virgil (p. 8), Cerda (p. 8), Tzetes (pp. 9,23), Cicero (p. 30), and Euripides (p. 35). Thus the manuscript notes on this single ode cite more works extraneous to the printed edition than we find in all of Milton's Greek marginalia.

Supplementation of the scholarly apparatus, then, is the primary purpose of the Pindar annotator. To the printed commentary he adds, among other things, cross references to other Pindar odes[20] and more precise references to works already cited in the notes.[21] He returns transliterations and scriptural quotations in Latin to the original Hebrew.[22] He records or summarizes in the margin the content of the commentary.[23] To the printed index he adds omitted items or citations,[24]


Page 82
and on his own initiative he compiles an "Index omnium authorum qui in opere citantur, exceptis Homero et Pindari Scholiaste, quos ubique." The content of the Pindar notes thus suggests a scholar working towards a new edition or a teacher perfecting his private copy for what he considered the better instruction of his pupils.

Such, then, are our reasons for questioning Milton's ownership and annotation of the Harvard Pindar. Nothing now in the volume, printed or written, indicates that Milton was associated with the volume before the 1871 sale catalogue. Pickering's argument that the notes are Milton's because the Pindar page references to Lycophron fit Milton's edition offers nothing conclusive. In fact, it provokes at least two questions that militate against Milton's authorship of the notes. If, as has been generally accepted, the Pindar entry on p. 756 (Δοξà τῷ θεῷ / Jun: 17 1630./ et Sept: 28. 1630.)[25] indicates the period of annotation, then how could Milton have referred to his copy of Lycophron, which he did not purchase until 1634? Or if the Pindar annotations belong to a period after he procured his Lycophron, then why do not the Pindar references to Euripides also fit the page numbers of Milton's copy, which he purchased in the same year as his Lycophron? Added to these difficulties are the differences in form and content revealed by comparison of the Pindar notes with the authentic Milton marginalia. The Pindar notes show different habits of annotation and handwriting and suggest scholarly interests other than those of Milton. We are therefore reluctantly forced to conclude that Milton's ownership of the Harvard Pindar has yet to be demonstrated, and that available evidence all points to a non-Miltonic origin for the annotations found therein.

If we are correct in this conclusion, then the results of much laborious work — particularly by the Columbia editors and Fletcher — will have to be discarded. The Pindar annotations should not be included in future editions of Milton's work. The nature of Milton's Greek studies at Cambridge and his use of Eustathius's Homer and Vulcanius's Callimachus will have to be established on evidence other than the Pindar volume. And scholars working with the Milton manuscripts may continue to date Milton's handwriting on the basis of the form of e that it contains; or at least, the Harvard Pindar should not be invoked to question the validity of that practice.

Plate 1

Page Plate 1

Plate 2

Page Plate 2

Plate 3

Page Plate 3

Plate 3 verso

Page Plate 3 verso



We quote the clipping from the catalogue now pasted inside the front cover of the volume, correcting one typographical error and making no attempt to reproduce the variations of type found in the original. The description—similarly presented—appears also in N&Q, 4th. Ser., VII (1871), 117.


A Milton Handbook (1947), p. 387.


XVIII, 276-304, 565-66.


Harris Francis Fletcher, The Intellectual Development of John Milton, II (1961), 281, 284-86.


Fletcher, "Milton's Homer," JEGP, XXXVIII (1939), 229-32; Nathan Dane II, "Milton's Callimachus," MLN, LVI (1941), 278-79.


Fletcher, Intellectual Development, II, 281-84.


Cited Plates II-III and footnotes as Pin., Dan., Ara., Lyc., Eur., Dio., and Her. To save space in Plate III, we omit the abbreviation for page. Thus, for instance, "Dio. 177" indicates Dio Chrysostom, p. 177.


See Plate III, Dio. 177, Ara. 52, Eur. II, 658, 620.


See Plate II, Pin. pp. 18, 601, 83, 12, 23.


See Plate III, Eur. II, 710; Her. 93.


See Plate II, Pin, pp. 437, 8, 201, 202, 5571, 602.


XVIII, 279, 14. We reproduce the misread note in Plate II, Pin. p. 12.


Cf. Plate II, Pin. pp. 603, 11, 9 and Plate III, Ara. 51; Eur. II, 430; Her. 337.


Cf. Plate II, Pin. pp. 23, 41, 9 and Plate III, Lyc. 21, 174.


Cf. Plate II, Pin. pp. 22, 10 and Plate III, Lyc. 63; Eur. II, 625, 616.


Cf. Plate II, Pin. pp. 5572, 756 and Plate III, Eur. II, 771, 427, 464. Those wishing to pursue differences into the Latin may compare the Pindar annotator's and Milton's writing of the initial consonant cluster of "Tzetzes", which the Pindar scribe (Plate II, pp. 83, 9) spells "Tzet-" and Milton (Plate III, Lyc. 174, 40) "Tset-". Another peculiarity not found in Milton's notes but frequent in the Pindar annotations is the use of "u" instead of "v". Instances of this practice, but not illustrated in our plates, are "auarus" (p. 2), "uocat" (p. 3), "uellem" (p. 8), et passim.


For a detailed analysis of the marginalia in this volume, see "Milton's Annotations of Aratus," PMLA, LXX (1955), 1090-1106.


The references to Scaliger in this volume (pp. 156, 163) derive from Canter's commentary, which is included in Milton's edition. A detailed study of Milton's Lycophron marginalia has yet to be published.


For an identification of the some 146 non-Miltonic entries and a general discussion of Milton's notes in this work, see "Milton's Annotations of Euripides," JEGP, LX (1961), 680-687.


See, for instance, "Nem: Od: 7 Str: 2" (p. 9), "Pyth: Od: 2 Str: 2." (p. 18), and similar notes on pp. 20, 23, 27, 40, 44, et passim.


See, for instance, "pag: 354" (p. 3), "pag: 709" (p. 79), and similar notes on pp. 118, 131, 191, 227, 228, et passim.


See notes on pp. 5, 137, 185, 386.


See, for instance, "AEgina populosa" (p. 151), "Archilochi hymnus" (p. 162), "Adjectiva in tribus generibus suis adverbiascunt" (p. 163), and similar notes on pp. 164, 215, 251, 253 et passim.


The great amount of labor involved in perfecting the printed index may be suggested by pointing out that under the letter A alone, the Pindar annotator supplies over 140 omissions.


We reproduce the entry in Plate II.