University of Virginia Library

Despite the present writer's private search and the more thorough efforts of the revisers of the Short-Title Catalogue (STC), no edition not previously recorded has been turned up, although more copies of all known editions except O1 and O4 have been found than had been noticed before. In this note the STC numbers are of the current first edition and do not reflect the revised numbering that will presumably be used in the new STC now in preparation. Since two editions are known only in single copies, it would not be surprising if some editions have been lost. However, one must work with what one has, and the following account assumes a contiguous relationship between all editions, an assumption that, in fact, is supported by the evidence of the texts.

This imprint, often a false one, was a precaution for some books that could not be licensed and were surreptitiously printed although, apparently, sometimes openly sold. The precaution was a wise one, for on 1 June 1599 the Stationers' Register records that the episcopal authorities called in and commanded to be burned, among others, 'Davyes Epigrams, with Marlowes Elegys'. The order listed has usually led to the conjecture that the book so suppressed would have been either the first or the second edition in which Davies' Epigrams precede the Elegies. It is further conjectured that the objection was more to the Davies than to the Marlowe section but the evidence is uncertain.

These were I.i,ii,iii,v,xiii,xv; II.iv,x; III. vi, xiii, the latter two being vii and xiv in modern editions of Ovid. III.v, not always present in early editions, was not translated. The order in O1-2 is I.i., I.iii, I.v, III.xiii, I.xv (misnumbered II.xv), I.xiii, II. iv, II.x, III.vi, I.ii.

I.xv is also printed in a translation by B. I., i.e. Ben Jonson, who wrote it for his Poetaster, acted in 1601 and printed in 1602.

The date 1595 is approved by J. M. Nosworthy, "The Publication of Marlowe's Elegies and Davies' Epigrams," R.E.S., 15 (1964), 397-398, but the evidence is suspect. An earlier date may be possible, as has been suggested by others on the basis of reference and quotation. However, the quotation in Nashe's Unfortunate Traveller (registered September, 1593) is of no evidential value since it comes from an elegy first printed in O3.

This Pforzheimer copy has been collated against photographs of the BM copy and found to be invariant.

For example, in Hamlet I.ii.198 (TLN 389) 'the dead vast and middle of the night' found in many modern texts, which comes from the Bad Quarto, could be entertained only with difficulty if the editor did not believe that in some manner Q2 'wast' corrupted the pure reading 'vast' in its copy and passed this misreading 'wast' on to the Folio. If F were set from an independent manuscript, instead of from annotated Q2, one could argue, its support of 'wast' would be weighty indeed unless it were conjectured to be an archetypal error.

The only O2 variants from O1 that O3 prints are I.ii.21 O1 needes] O2-3 needst (O3 need'st); I.ii.36 O1 hath] O2-3 have; I.xiii.47 O1 chid] O2-3 chide; III.xiii.31 O1 makes] O2-3 make. Actually, the copy for O3 as a whole is a complicated one and the statement that O1 was the printer's copy applies only to the Elegies. For a discussion, see below.

On the other hand, the Latin would permit either plural or singular, nec meus innumeris renovatur campus aratis, which the Loeb Classical Library edition translates as 'and my fields are not renewed with ploughshares numberless'.

O3 prints lines 22-26 as:

On the other hand, the error is less explicable from the O2 setting: Huge Okes, hard Adamãts might she haue moued & with sweet words, cause deaf rockes to haue lo- Worthy she was to moue both gods & men (ued

Contributing no doubt to the error is the fact that in O1-2 names are printed in roman although consistently placed in italic by O3, but not 'Love' for Cupid or Venus. The O3 compositor seeing 'Loue knowes with such like praiers I dayly moue him' in his copy, especially with 'Loue' not wholly clear because of a damaged 'L', could well expect that the poet was swearing by Jove, not Cytherea (as in the Latin). I am indebted to Mr. Kendon Stubbs for privately communicating to me the case of the damaged 'n' in O1 I.v.5 which barely prints in CSmH 'sunne' and might just possibly explain the misprint 'Suune' in O3, unless this is simply a fortuitous turned 'n'.

The Latin here is Quaea mihi ventura est, siquidem ventura, senectus, | cum desit numeris ipsa iuventa suis?

It might be useful in tracking him down to know that he did not have in his cases a ligatured long-s with k and set this combination always with two sorts. The printer of O6 had few roman 'k' sorts and frequently had to resort to italic.

Since O3 customarily mends metrical deficiencies, real or fancied, both in the Epigrams and in the Elegies, it is unusual to find it reprinting this faulty line without change. But fortunately for the case in hand it was preserved, perhaps because no alternative suggested itself to the editor. If so, in this instance he did not consult the Latin.

This is an example of O3's tendency to alter any rough metre.

Type-shortages show that some sheets, like B, were set by formes from cast-off copy. The possible oddity I have noticed is that the general trend in sheets A and B to the spellings 'mistresse' or 'mistrisse' begins toward the latter part of sheet C to give place to 'mistris', which had not appeared before with any frequency. The punctuation, also, grows somewhat heavier. Other characteristics do not seem to change, however, at least so far as I have observed. For example, the invariable spelling 'doest' (only one exception) is preserved throughout, and the italicized treatment of names is constant. The unusual hyphenation persists: see 'out-stander' (III. v.27) on sig. E2v, or 'nimph-Neœra' in the next line, 'hood-winckt' (III.v.79) on sig. E3v, 'fal-shood' (III.viii.30) on sig. E7, 'in-tombe' (III.xiii.23) on sig. F3.

To identify the editor of the Epigrams with the editor of the Elegies is not necessarily to require him to have annotated the ten Certain Elegies in O1 printed copy for the O3 compositor. If, instead, it was the printer himself who chose to perform the annotation in order to speed up composition — or merely consulted O1 while typesetting — one would need to take it that the O3 variants in the Elegies were present in the manuscript, whether originally or by alteration of the editor who changed the Epigram readings. The identity of this person is, of course, a mystery. Whether it was a literary-minded publisher, whether the publisher hired some scholar or literary man to oversee and bring up to date an old-fashioned work now being given an important new complete edition, or whether they were already present in the manuscript, can be a subject only for speculation.

Just possibly the word 'flying' was not entirely legible in the print. In the BM copy the inking is poor for some letters of this word.

In I.xv the lines are 'The world shall of Callimachus euer speake, | His Arte excell'd, although his witte was weake.' In II.iv they are 'I thinke what one vndeckt would be, being drest | Is she attir'd, then shew her graces best. | A white wench thralles me, so doth golden yellowe | And nut-browne girles in doing haue no fellowe'. No reason exists to doubt that these lines are Marlowe's own.

O3 corrects seven O1 misreadings and misprints, such as O3 'What' for O1 'That' (I.i.11), 'sunne' for O1 'sonne' (I.i.15), or 'number' for 'numbers' (I.i.22). It makes nine misprints or misreadings of its own, such as 'slackt' for O1 'shakt' (L.ii.12), 'tride' for O1 'tyrde' (I.v.25), 'chide' for O1 'chid' (I.xiii.47), or 'sire' for O1 'sir' (III.vi.11). These are not always easy to separate from such O1 errors corrected by O3 as 'workes' for O1 'worke' (I.i.21), 'Loue' for O1 'I' (I.i.22), 'Am' for O1 'And' (II.iv.8), or 'glance' for O1 'glasse', which total seven, or the four O3 errors where O1 was correct such as O3 'Temple' for O1 'tempe' (I.i.19), 'Ioue' for O1 'Loue' (I.iii. 4), 'rustie' for O1 'dustie' (I.xv.4) or the omission of 'now' (III.vi.70).

These are the O3 omission of O1 'both do' (I.xiii.21), O3 'And this' for O1 'This' (II.x.8), O3 'not' for O1 'her not' (III.vi.3), and O3 omission of O1 'and' (III.vi.47).

These modernizations remove several examples of the attributive case such as O3 'needles points' for O1 'needle poynts' (III.vi.30), or — better — 'nights pranckes' for O1 'night prankes' (III.xiii.7); create concord between verb and noun, as 'life . . . giues' for O1 'life . . . giue' (I.iii.13) or 'reasons make' for O1 'reasons makes' (II.iv.10); or mend grammar such as the objective case 'her I like' for O1 'she I like' (II.vi.29).

These four cases lose any possible significance on examination. One group is quite definitely O3 sophistication (the plural 'souldiours' and the change in pronoun to 'their' in O3 for O1 and Latin singular and 'his' in II.x.31-32, which make for rough syntax and lack of parallelism). In I.xv.4 O3 'rustie' for O1 and Latin 'dustie' is almost certainly a sophistication as well, if not a memorial error. In the third case the motive for the O3 variant 'wooers' of Lais for O1 'louers' (I.v.12), the Latin being et multis Lais amata viris, is not clear but the change may be sophistication. The fourth is O3 'ore' for O1 'on the sea' (I.xiii.1), fairly clearly a sophistication (super oceanum venit).

Perhaps in this category, although the point is a fine one owing to the ambiguity of the O1 modification, is the O3 straightening-out with 'And she thats coy I like for being no clowne, | Me thinkes she would be nimble when shees downe' of O1 'And she thats coy I like, for being no clowne, | Me thinkes she should be nimble when shees downe' (II.iv.13-14), the Latin being sive procax aliqua est, capior, quia rustica non est, spemque dat in molli esse toro. The O3 version is closer to spemque, and clarifies the O1 comma after 'like' (perhaps no more than a conventional caesural punctuation pause) which obscures the modification by making it possible for the 'for' to be taken as a conjunction that would associate her nimbleness directly with her lack of rusticity, a notion not encouraged by the Latin.

Another very literal correction but perhaps dictated as much by smoothing the language as by accuracy is O3 'Her armes farre whiter then the Sythian snow' for O1 'That were as white as is the Scithean snow' (III.vi.8). The Latin is bracchia Sithonia candidiora nive. In III.vi.18 O3 'When in my prime my force is spent and gone' is closer to cum desit numeris ipsa iuventa suis than O1 'Seeing in my prime'; but it is more likely that the editor made this change to reduce the line from eleven to ten syllables than to alter the translation of cum, for he keeps the same Marlovian use of 'seeing' in line 70 when the line totals ten syllables exactly. A real correction from the Latin (but perhaps from the Jonson version) comes in O3's 'Whilst Rome of all the conquered world is head' for O1 'conquering world', perhaps a mistranslation of Roma triumphati dum caput vrbis erit, or perhaps a scribal error. If the O3 compositor were merely consulting O1, many of the relatively indifferent words may have come from the O3 manuscript and need not reflect the activity of an annotating editor. This is an attractive hypothesis.

For instance, see O3 'why made King to refuse it?' in III.vi.49, which tightens O1 'why made king? and refusde it' in the direction of quo regna sine usu.

The Latin is in freta collectas alta quid addis aquas. This O3 variant, then, does not even have the excuse of a more literal rendition of the Latin such as seems to be behind the syntactical rearrangement in O3, I.xv.31, 'To verse let Kings giue place, and Kingly showes' (cedant carminibus reges regumque triumphi) of O1 'Let Kings giue place to verse and kingly showes', which is itself closer to Jonson's line 'Kings shall giue place to it, and Kingly showes'. But it seems clear that Jonson knew Marlowe's translation when he came to pen his own version.

O1 'yeelding or striuing [O3: strugling] doe we giue him might' for Cedimus, an subitum luctando accendimus ignem. Perhaps again O3 is merely modernizing the language.

These missing lines in O1 present an insoluble problem. The four lines in II.iv, particularly, are of a nature that would lead Marlowe to want to translate them. Whether in the manuscript given to the printer or somehow in the printing, however, they were skipped, for they must be thought to have been present in Marlowe's papers. At first sight a change on O3 from the O1 version in II.iv immediately following the four-line gap might seem to have something to do with a revision consequent upon their restoration. That is, after the couplet on Leda, O1 continues, 'Yellow trest is shee', but in O3 the reading is 'Amber trest'. One could speculate idly that 'Amber' was substituted in O3 by an author who was varying it in order to avoid the repetition of 'yellow' in the third of the missing lines — 'A white wench thralles me, so doth golden yellowe'. But the change probably has nothing to do with the omission since it aligns itself with other examples of editorial exactness. The phrase 'so doth golden yellowe' translates capiet me flava puella. But whereas O1 'Yellow trest is shee, then on the morne thinke I' also translates at the start seu flavent in identical terms as 'yellow', the continuation of the translation omits a distinction that follows, seu flavent, placuit croceis Auroris capillis. It is the croceis or saffron-yellow color that is being transferred in O3 from Aurora to the yellow-tressed girl. In passing one may notice the oddity that the first of the four lines restored in O3 has just the same sort of mistaken punctuation distorting the sense that O3 had sometimes followed from O1 and sometimes corrected:

Is she attir'd, then shew her graces best. Here the comma after 'be' is acceptable only if another comma or stronger stop is added after 'drest' to prevent a run-on line, and a question mark though not required would not be amiss after 'attird'. The Latin reads non est culta—subit, quid cultae accedere possit; | ornata est—dotes exhibet ipsa suas.

One example of possible consultation of the manuscript must be remarked although it may be thought a doubtful one. This is I.xv.39 which reads in O1:

O3, however, reads 'rakes my bones', which is argued for by the Methuen editor as meaning 'covers (sc. under the materials of the fire which have been raked together.' It is true that O.E.D. quotes some examples of this meaning under vb.II.4 'To cover with, or bury under, something brought together with, or as with a rake' and in 5. spec. gives quotations to illustrate 'To cover (a fire) with ashes or small coal in order to keep it in without active burning', the best of which is from Brathwaite's Strappado (1615), 'Yet shall not . . . those accomplisht parts . . . Lie rak't in Ashes.' The Methuen editor remarks another quotation with 'up' from 5.b, Stapylton's Strada's Low-Country Warres (1650), 'His Indignation, then rak'd up in Embers, would in time breake out.' The Latin reads, ergo etiam cum me supremus adederit ignis, in which the verb adedo means 'to eat away, gnaw at', and, specifically of fire, 'consume'. The contrast is then between what happens to his bones in the funeral fire, and 'Ile liue' (vivam). (Jonson's version reads 'when this body falls in funeral fire'.) It is possible that Marlowe is thinking of Death with a rake covering up his bones in the funeral fire so that they will be utterly consumed. This is very likely the intention, although the subsequent editions did not understand it. On the other hand, one can worry a bit about the connection between the fire gnawing away at his bones (adederit ignis) and Death torturing them by the fire as if tearing them apart on the rack ('to pull or tear [them] apart, to separate by force, to break up' O.E.D. vb.2b obs.). But this is perhaps fanciful, and the O3 variant 'rakes' may have come from manuscript. However, it is worth notice that 'rackes' as in O1 is a rare spelling of 'rakes'; so, in the end, it is quite possible all the O3 editor, or compositor, was doing was modernizing the spelling. The O3 form could not have come from the Latin, which does not suggest the word at all.

This care for the Latin readings did not prevent some slips, however, such as his passing 'thee conquered' in the last line of I.ii instead of altering 'thee' to 'the', or his leaving untouched 'fathers hoord' (I.xv. 17) and other common errors that have been listed. There are a scattering of examples of Marlowe's mistranslations being left unmended, such as II.x.19 'soft loue' for saevus (cruel), which Marlowe sems to have misread as suavis. Or I.iii.18, where O3's modernization of O1 'or' to 'ere' in 'Ile liue with thee, and die, or thou shalt grieue' does not mend the sense to conform to vivere contingat teque dolente mori (and to be sorrowed over by you when I die). Various of these examples are complicated, however, either by the difficulty of altering Marlowe except in single words or by the uncertainty as to the readings in the edition used by the editor.

Actually, the signing only of the first three elegies seems to reflect the copy, and the cessation to represent a compositorial decision. The main point is that they start off being signed. The first three are contained in sheet E, and the system without signing begins with sheet F. It is possible that some break took place between these two sheets on the evidence of the treatment of the heading capitals which in sheet E take up two lines of indented text but beginning with sheet F three lines, although the size remains the same. Yet no clear sign of a change of compositors can be seen and the spelling system remains the same, except for the anomalous use of 'hir' in I.iii on sig. E3 but 'her' elsewhere in E and invariably in sheets F-G. Moreover, type shortages in sheet E that dictated the substitution of 'VV' or 'w' for 'W', of italic 'F' for 'F', and of italic 'T' for 'T', continue in sheet F, just as they had appeared in sheets A, B, and C. The reason for the typographical change is obscure although it may have had an effect on the omission of the signatures. But at least the single compositor seems to remain constant.

A hypothesis that O1 was set from a portion of the manuscript behind O3 and then returned cannot be seriously entertained.

Other copies may be found in the Dyce collection of the Victoria and Albert Museum, in the Bodleian, Huntington, Folger (3 copies), Pierpont Morgan, Carl H. Pforzheimer, Harvard, Yale, and University of Illinois Libraries.

Other copies may be found in the Bodleian, Huntington, Folger, and Newberry Libraries, and in the libraries of the University of Hull and Harvard and Yale.