University of Virginia Library

A Bibliographical Study of Parthenissa by Roger Boyle Earl of Orrery
by
C. William Miller

I.

THE EARLIEST PUBLICATION OF THE POPULAR heroic dramatist, Roger Boyle, Earl of Orrery, is his prose romance Parthenissa. Despite its literary barrenness, almost every historian of seventeenth-century English literature has felt obliged to mention the work because it stands as the only serious lengthy imitation in English of the French heroico-historical romances written by Gomberville, La Calprenède, and Scudéry. Some of the bibliographical problems which this English romance presents have, on the other hand, caught the attention of a few scholars, but no one has made a really intensive study of the various editions and issues of the romance as a problem in itself.

The order of publication of certain basic editions of Parthenissa is above dispute, and since it is generally agreed upon by William Carew Hazlitt,[1] Arundell Esdaile,[2] and Donald G.

Wing,[3] the three bibliographers of the romance, its presentation at this point will serve to assist the reader in following later explanations. At the end of this paper I shall present a more complete list incorporating several additional items cited in the course of the discussion.

• (1) Parthenissa (Part I, Books 1-6).[4] Printed for Richard Lownes, London, 1654. Collation: 4°, A-2L4 2M2. A1: title-page, v. blank.
• (2) Parthenissa In Four Parts. Printed for Henry Herringman, London, 1655. The First Part Collation: 4°, A4(±A1; A3+A2+B2) B-3G4 [3H]1. A1: title-page, v. blank. The Second Part Collation: 4°, i1 A-3M4 3N1. i1: title-page, v. blank. The Third Part Collation: 4°, i1 *2 A-3G4 3H2. i1: title-page, v. blank. The Fourth Part Collation: 4°, i1 A-3E4. i1: title-page, v. blank.
• (3) Parthenissa The Fifth Part. Printed by T. R. and E. M. for Henry Herringman, London, 1656. Collation: 4°, [A]1 B-2Q4 2R1. [A]1: title-page, v. blank.

---

117




• (4) Parthenissa The Sixth Part. Printed for Henry Herringman, London, 1669. Collation: 4°, A-2Q4. A1: title-page, v. blank.
• (5) Parthenissa The Six Volumes Compleat. Printed by T. N. for Henry Herringman, London, 1676. Collation: 2°, A2 B-3E4 3F2, 3Q-5K4 (5K4 blank(?), lacking in observed copies). A1: title-page, v. blank.

The edition of Parthenissa which Esdaile prefixed to the above list and which he and all literary historians after him have considered to be the first,[5] he records thus:

To this description Esdaile appended the note: "There is a transcript of the Waterford Title-pages by H. Bradshaw in U[niversity] L[ibrary] C[ambridge]." Hence Esdaile, who otherwise gave a location for each book he listed, had not seen a copy of the romance with Waterford imprints, but in making his entry, he believed at the time, as he has written me, that a copy with these title-pages must have existed somewhere.[7]

The "transcript" to which he referred is set down in Bradshaw's handwriting on four separate leaves inserted after each of the four title-pages of copies of the four parts of Herringman's Parthenissa (London, 1655), which Bradshaw once owned and later gave to University Library Cambridge along with his extensive collection of Irish books.[8]

In these notes Bradshaw makes no mention of the owner or the library location of the Waterford set, the titles of which he is presumably transcribing in his library manuscript insertions. Yet such facts as these he might perhaps be expected to include in a detailed description of a rare example of provincial Irish printing not in his private collection or in that of University Library Cambridge. In fact, Bradshaw never says that he is making notes from a particular set of volumes. On the verso of each of the four inserted leaves he simply summarizes the contents, including story-headings and pagination, of each of the four volumes, which except for the title-pages are identical with those in the Herringman, London, 1655 set. On the recto of each leaf, however, he seems to be describing a series of title-pages, all recorded with the same careful regard for detail as this first one:

To assume, as Esdaile did, that Bradshaw's notes constitute transcripts of actual printed title-pages, a fact presumably indicated by the lineation and the sketches of ornaments in the manuscript insertions, obliges the bibliographer to face at least one inconsistency in the practice of either Boyle or de Pienne. Boyle notes in his preface, inserted in the first gathering of Parthenissa, The First Part, Herringman, 1655, and written after 'the finishing of the Fourth Tome,' or Herringman's 'The Fourth Part,' that while he had originally planned to write Parthenissa in six tomes he has so far completed only four and that if he undertakes the last two it will be in penance for having done the first four. In fact, it was not until 1669, some fifteen years later, that Boyle completed the sixth 'tome' and

had it printed and published through Herringman in London. Yet Bradshaw's notes carry 'In six Tomes' in the description of all four Waterford title-pages presumably printed in 1654-55.

A much more plausible alternative interpretation of the Bradshaw notes, which can be supported by abundant evidence and which accounts for the apparent inconsistency just presented, is that Bradshaw shrewdly recognizing the sheets of the romance Parthenissa in quarto with Herringman, London, 1655, prefixed title-pages to be actually the presswork of an Irish provincial printer, undertook to construct a set of hypothetical title-pages which would describe as accurately as he could determine them the facts of the printing of the romance. This explanation of the objectives of Bradshaw's investigation certainly squares with the "natural-history" method which he first applied with great success to his historical classification of volumes lacking dates and printers' names in University Library Cambridge.[9]

Strong evidence supporting the "construction" explanation exists in the form of personal testimonies from G. W. Prothero, Bradshaw's biographer; Francis Jenkinson, his bibliographical protégé and successor as librarian of U.L.C.; and Sir Norman Moore, one of his friends. Prothero writes:

Jenkinson, after examining Esdaile's newly published List (1912), wrote the following in a private letter to the compiler: Sir Shane Leslie, in a comparatively recent essay on Bradshaw, substantiates the opinions of Prothero and Jenkinson with information given him by one of Bradshaw's friends.

Clearly these accounts are at odds on how Bradshaw proved his case, but all agree on the central point that he was attempting to establish the actual printer, place of publication, and date of Parthenissa from available bibliographical information, not that he was recording information derived from the titles of any copies which he had seen.

There now remains, in substantiation of the personal testimony, the necessity for demonstrating how Bradshaw probably went about constructing his hypothetical title-pages. He had available to him in his own collection the entire six parts of the romance in quarto issued by Herringman, London, 1655-69, the Herringman 1676 folio, and the two issues of Cook's Monarchy No Creature of God's Making, printed and sold in Waterford by de Pienne (1651) and reissued in London by Brewster

in 1652.[14] His biographer states that the Parthenissa problem was one of the last on which he had been working before his death in 1886; therefore in all probability the new light on the Parthenissa problem which gave impetus to his research at that late date was the entry "Parthenissa 4° 1654" in the second volume of the British Museum Catalogue published in 1884. The only scrap of evidence of Bradshaw's work on the romance, other than the title-page notes, is the description of that B. M. copy of the R. Lownes 1654 London edition—the one preceding Herringman's—preserved in his notebooks.[15] It is improbable that he was aided by manuscript copies of either the Stationers' Register (1913-14) or The Letters of Dorothy Osborne (1888), both published after his death.

Of the facts included in the Bradshaw title-page notes for the first tome[16] quoted earlier, the title of the romance, the dedication, and the part and book numbers could have been readily derived from the Herringman 1655 title-pages. 'Parthenissa A Romance' follows the Herringman title-page verbatim. In place of Herringman's 'In Four Parts', Bradshaw has substituted 'In six Tomes', a phrase suggested by the words 'The Last Part | The Sixth Tome' on the title-page of the Herringman 1669 quarto edition of The Sixth Part. The words 'Composed by the Right Honble | The Lord Broghill' follow the Lownes title-page phrasing exactly, even to capitalization, punctuation, lineation, and abbreviation except that Lownes prints 'Honoble' and Bradshaw writes 'Honble'. Having previously referred to the six major divisions of the romance as 'tomes,' Bradshaw is only consistent in substituting 'The First Tome' in place of Herringman's 'The First Part,' while the

phrase 'containing the first four Books of The First Part,' which appears on neither the Lownes nor the Herringman title-pages, is certainly Bradshaw's attempt to define precisely the contents of each 'tome' in the face of the confusing dual system of numbering the divisions of the romance found in the Herringman publication. The words 'Dedicated to the Lady Northumberland' appear on both the Lownes and Herringman title-pages. The imprint which Bradshaw composed follows, except for two deletions,[17] the phrasing employed by de Pienne in his imprint of Cook's Monarchy No Creature of God's Making.

Bradshaw's identification of de Pienne as the printer and Waterford as the place of printing resulted almost certainly from his intimate knowledge of the history and press work of early Irish printers, which enabled him to identify from memory the characteristics of de Pienne's type, border designs, and craftsmanship in the Herringman quartos. If Bradshaw had had any doubts about the accuracy of his identification, even a casual comparison of his copy of the Herringman Parthenissa quartos with his copies of Cook's Monarchy printed by de Pienne in 1651 would have revealed to him that the fonts, border designs, and decorative capitals were identical. Further, de Pienne's faulty workmanship in permitting the 'beards' of the letters to print[18] on the top and bottom lines of most of the pages of Parthenissa is also evident on almost every page of Cook's book. De Pienne's printing career centered on the towns of Cork and Waterford; the two extant books printed with his name on the title-page in the 1650's bear Waterford imprints.[19]

Bradshaw's settling on the dates 1654-55 for the printing of the romance is incorrect, as I shall demonstrate later in this paper. Sir Shane Leslie's statement that Bradshaw "proved from

the types that it [Part One] was printed . . . in the year 1654' is obviously inaccurate since de Pienne used the same types in 1651. Equally questionable is Prothero's statement that 'Bradshaw ingeniously proved [a part of Parthenissa] to have been printed in 1654 at Waterford where the author was at that time staying.' An author's residence in a particular town during a certain year is hardly proof for establishing the printing date of a book, especially when the Boyle family seat, Lismore Castle, was located only a short distance from Waterford. The fact is that by June, 1654, Boyle had left Ireland for London.[20] If Bradshaw had proved to his satisfaction that Boyle was residing in Waterford during the first half of that year, he did so on the basis of evidence unknown to Boyle's careful modern biographer, Professor W. S. Clark, II, who had access to all extant family papers. In all probability Bradshaw settled on the dates 1654-55 by reasoning that de Pienne working alone or at most with one assistant must have set up and printed off approximately half of the text of the lengthy romance sometime in 1654 if he was able to have completed and shipped to London the sheets of all four volumes in time for Herringman to publish them with his own titles in 1655. Or he might have conjectured, or perhaps even determined by collation, that the Lownes 1654 Parthenissa, Part I, Books 1-6, was set from the de Pienne text and hence that the bulk of the first two tomes must have been printed off by 1654.

The only detail of Bradshaw's title-page notes now unaccounted for is his ornament sketches. All four ornaments are drawn in the meagerest detail and do not appear to be careful efforts to copy particular title-page ornaments. They most nearly resemble large asterisks; the two on the title-pages of the tomes which Bradshaw believed were printed in 1654 are identical and are very slightly differentiated from the two identical ones on the title-pages of the tomes dated 1655. De

Pienne did not use an ornament on the title-page of either his Cork Eikon Basilike or his Waterford Monarchy, nor did Herringman's printer use an ornament on his London 1655 title-pages. Lownes, however, did use one on his 1654 title-page, and since it has been pointed out that Bradshaw depended on the Lownes title-page for the lines of his hypothetically constructed title-pages stating the authorship, even to the point of reproducing capitalization, lineation, punctuation, and abbreviation, one can only suppose that in drawing up his notes in title-page form he satisfied a decorative whim[21] by inserting an ornament as Lownes had done, making the ornaments for the title-pages of Tomes One and Two, dated 1654, identical and those for the title-pages of Tomes Three and Four, dated 1655, likewise identical.

The evidence presented so far, therefore, establishes two points: (1) Mr. Esdaile was mistaken in interpreting Bradshaw's title-page reconstructions as recordings of the titles of an actual edition printed in 1654-55, and (2) Bradshaw's significant contribution to Parthenissa bibliography was his identifying de Pienne as the printer of the quarto sheets offered for sale in London by Herringman in 1655.

2.

Earlier in this paper the printing dates 1654-55 set down by Bradshaw in his reconstructions were called inaccurate. It remains now to prove the validity of that assertion and to establish the actual date of printing.

As early as September, 1653, Dorothy Osborne, in writing to William Temple on the subject of romances, states: 'My Lord Broghill sure will give us something worth the reading.'[22]

On October 12, 1653, Herringman had Parthenissa entered in the Stationers' Register as 'a romance heretofore written by the Lo: Broghall.'[23] In an issue of the Mercurius Politicus dated 'From Thursday January 19. to Thursday January 26. 1654' the Lownes Parthenissa was advertised for sale. By mid-February, 1653/54, Dorothy Osborne had received a copy of Parthenissa from her brother in London and had almost finished reading it. These pieces of evidence attest to the fact that Boyle had certainly written at least six books of Part One, the whole of the contents of the Lownes Parthenissa, sometime before 1654. If it now can be established that the Lownes text was set from that of de Pienne—and it is reasonable to suppose that a second edition would be set from printed copy rather than from manuscript provided that a printed copy were available—then de Pienne must have printed a portion of Part One sometime before 1654 also.

The collation of the text of 'The First Part' (Part One, Books 1-4), which Bradshaw had identified as the press work of de Pienne published by Herringman, with the corresponding text of the first four books of Part One printed for Lownes reveals three types of evidence pointing to Lownes' dependence on the de Pienne text.

(1) In eleven instances de Pienne undertook to enclose a phrase in parentheses and failed to include either the opening or closing parenthesis. The Lownes compositor erred in identical fashion on nine occasions, managing to supply de Pienne's omissions only twice.

(2) De Pienne almost habitually used a two-line-high capital for the first word in the text following either the close of an intercalated episode or the presentation of an epistle written by one romance character to another. Seventeen times de Pienne used a two-line-high capital, and eight times he employed a capital of the same font as the rest of the line. In all twenty-five instances the Lownes compositor followed the

de Pienne practice. On four occasions de Pienne introduced a new division of text with a head decoration and a large capital; on each occasion the Lownes compositor did the same. Three times de Pienne set the first letter of the first word in a new segment of text with a two-line-high capital and then set the rest of the letters in that word in one-line-high capitals. The Lownes compositor in each instance set the first word in like manner.

(3) The Lownes compositor, after correcting repeatedly in his text the literal errors committed by de Pienne, finally permitted inadvertently the following error to creep into his London text. The passage which in de Pienne reads: '. . . for Canitius and Castus with those 400 men that were yet in oae of the Groves for our Reserve', is reproduced in Lownes with the word one misprinted in the same odd way.[24]

The evidence, therefore, in favor of resetting is so conclusive that one may state definitely that the Lownes compositor set his text from de Pienne's, not from an independent manuscript, and hence that at least the first six books of Part One—all that was reprinted by Lownes—were printed by de Pienne at some time earlier than 1654, the date fixed by Bradshaw.

Thus the first of Bradshaw's conjectured dates is proved to be inaccurate, but, if the reader recalls, Bradshaw suggested the printing date 1654 for only the first and second tomes (Part I, Books 1-8). Tomes Three and Four (Part II, Books 1-8), he believed, were printed in 1655. External evidence to prove or disprove the validity of Bradshaw's second date is lacking, but the internal evidence of the headlines in the text of the four tomes printed by de Pienne and published by Herringman points to the fact that Bradshaw was also inaccurate in advancing the date 1655 and in believing that a break occurred between the printing of the first two and the last two quarto volumes.

De Pienne, in printing the four tomes of Parthenissa which Herringman published in London, used throughout two skeletons per sheet. By means of the inevitable alterations which from time to time creep into headlines[25] it has been possible to trace overlapping series starting with the inner forme of the Ee gathering in the first tome and ending with the outer forme of the last complete gathering in the fourth tome. It can be stated confidently, therefore, that there is no indication of any break in the printing of sufficient extent to lead to the distribution of the skeletons; instead, it may be concluded that printing was normally continuous.

The evidence gathered from the collation of the de Pienne and the Lownes texts proved that the Lownes text was reset from de Pienne's and hence that de Pienne's printing date for the first tome (Part I, Books 1-4) and half of the second tome (Part I, Books 5-6) was earlier than 1654. The evidence of the de Pienne headlines in all four tomes proves that printing was continuous. Hence one is obliged to conclude that not just Part One, Books 1-6, but actually all eight books of Part One and of Part Two, the whole of the four quarto volumes, were printed earlier than 1654.

How much earlier than 1654 de Pienne started the printing is the question that must next be considered. To give a definite answer would be well nigh impossible were it not for the existence of a copy of Parthenissa containing the de Pienne sheets as later published by Herringman but with a unique title-page dated 1651[26] as the first leaf, the text including the whole of Parts One and Two or the equivalent of the four "Parts" of the Herringman 1655 publication. This lone title-page in a copy found among the holdings of the University of Texas Library bears only the romance title "PARTHENISSA" in large

capitals and an imprint: "Printed in the yeare 1651." The type used to print this title-page is identical with that used elsewhere by de Pienne, and the leaf bears part of a watermark identical with that found on several other leaves in Part One. Because the set has been tightly bound into two volumes by a modern binder, it is impossible to determine positively whether the title-page A1 is conjugate with A4, the fourth leaf of the first gathering in fours, but the chain lines and complementary parts of the same watermark seem to meet, and the signatures of the other leaves seem to warrant such a conclusion.[27] The make-up of the first leaf makes it questionable whether de Pienne ever meant it to serve as a commercial title-page since the facts of authorship, printer, and place of printing normally included on his title-pages are missing. The presence of an imprint date rules out one's considering it a half-title, and the leaf does not bear the information found normally on a section-title. Its contents seem to resemble those intended simply for an identifying title-page on a work privately printed, the printer's expectation being that once the romance was completely printed and publication arrangements made, de Pienne or some bookseller would cancel the temporary title-page and substitute a normal one.

The bibliographical evidence offered earlier points definitely to de Pienne's printing the four quarto tomes of Parthenissa before 1654. The imprint on the title of the Texas copy fixes the date of printing as 1651. The evidence next to be presented suggests the likelihood of de Pienne's doing the printing in about the year 1651.

Professor W. S. Clark, II, Boyle's biographer, states that Boyle served as an officer in the Irish Commonwealth forces from April, 1650, to June, 1654, and that he was most busily engaged in campaigns during the years 1650-52.[28] Hence, in a

letter to me, Professor Clark stated that there might be a question whether Boyle could have found time to compose any large portion of his romance in the early 'fifties. The reasonable alternative explanation is, then, that Boyle must have completed the bulk of his writing sometime earlier. In any case all the campaigns in which Boyle served took place in the south of Ireland[29]; therefore, it is probable that he had a few opportunities to visit Lismore Castle, the family seat, and possibly at least one opportunity to make arrangements for the printing of his romance with de Pienne in Waterford, some thirty miles away.

Like Boyle, de Pienne had previously been royalist in his sympathies as his printing of the Eikon Basilike at Cork in 1649 attests, but by 1651 he had moved to Waterford and had begun operating the Commonwealth press from which he issued Cook's Monarchy No Creature of God's Making (1651) and An Act for the Settlement of Ireland (1652). According to an order issued by the Council for the Affairs of Ireland dated at Kilkenny on September 30, 1652, de Pienne's press was locked up and his salary stopped,[30] but at some point in the year 1651 he could certainly have initiated a private printing job for an officer high in the regard of the Commonwealth government by whom de Pienne himself was employed. Further, the fact that no books bearing de Pienne's imprint dated later than 1652 are known affords us at least the negative evidence for believing that de Pienne ceased printing in that year.

Establishing the date 1651 as in all probability correct for the printing of Parthenissa clears the way for a consideration of the complete setting and printing of this unique quarto in relation to its subsequent publication in London. The fact that it included all eight books of Part One and of Part Two, which Herringman offered for sale 'In Four Parts', each of Herringman's parts bearing a title-page, and yet that it exists with only an elementary title points to two conclusions: (1) De Pienne did

not offer the romance for sale in Ireland; otherwise he would have cancelled the elementary title-leaf and supplied titles bearing his own imprint for all four tomes. (2) The state of this set is that of the sheets as they came to Herringman, who cut out the first leaf, leaving the telltale stub on A4,[31] inserted his cancel title-leaf, and then supplied the three other divisions, which had never had title-leaves, with his own title-pages.

The University of Texas copy is, therefore, I suspect, the only extant copy of several perhaps retained in Ireland for distribution among Boyle's friends. Barring short printing of Herringman's title-pages, it is unlikely that the bookseller would have parted with a copy of the romance in London lacking his title-pages unless it were to the importunate author himself.

If this explanation of the state of the Texas University copy and of the de Pienne-Herringman relationship is accurate, the various editions of Parthenissa follow this order: (1) Parthenissa, Part One, Books 1-8, and Part Two, Books 1-8, 'Printed in the yeare 1651', constitutes the first issue of the first edition. (2) Parthenissa, Part One, Books 1-6, 'For R. Lownes . . . London, 1654', stands as a second setting of type done from the de Pienne text and is therefore a second edition, although it is presumably the first text offered for public sale. (3) Parthenissa, Part One, Books 1-8, and Part Two, Books 1-8, 'For H. Herringman . . . London, 1655', is the second issue of the first edition. Thus the second issue of the first edition of Parthenissa, Parts I and II, Books 1-8, was offered for sale after that of the Lownes second edition of Part One, Books 1-6. In addition, Parthenissa, Part One, Books 7-8, and all eight books of Part Two, which Herringman divided in two and called the Third and Fourth Parts, were offered for sale by Herringman presumably for the first time.

3.

Once Herringman had received the sheets from de Pienne, he did not, however, attempt to sell all of them at his own London shop, for one incomplete set of the romance, consisting only of "The First Part," "The Second Part," and "The Third Part," now in the Huntington Library,[32] exists with Herringman title-pages bearing the press-variant imprints of Humphrey Moseley,[33] the leading London bookseller and foremost dealer in romances at that time. How the title-pages for this Herringman-Moseley London publication of Parthenissa were printed is bibliographically informative, especially in view of an error on the printer's part which caused the title-page imprint for "The Third Part" to exist with three variants.

The London title-pages of the four parts of the romance all have the same general appearance, and all are printed in two colors of ink, black and red. The distribution of matter printed in the two colors of ink on the title-pages of all four parts follows this of the Herringman state of the 'The First Part':

A comparison of the four title-pages reveals at once that the title-pages for 'The First Part' and 'The Third Part' are printed with the exception of the necessary substitution of 'Third' for

'First' from one and the same setting of type and that the title-pages for 'The Second Part' and 'The Fourth Part' are likewise printed from one and the same setting of type, but a setting of type different from that used for the printing of the title-pages for the first and third parts.[34] Further, one of the title-pages bears half of a watermark on the outer edge of the leaf.[35] With the help of these clues the method of printing the title-pages may be readily reconstructed.

The printing was evidently done by a modified form of half-sheet imposition. The two different type-settings for Parts One and Two were imposed side by side in half of a regular quarto forme and printed together at one time. Upon completion of the run, the formes were unlocked, the part numbering was altered to 'Third' and 'Fourth,' and the title-pages for these parts were machined. Whether this second machining perfected the previous sheets or whether a complete run of printing and perfecting was made for each set of title-pages is impossible to determine.

Thus the general method of printing may be explained, but there are two other printing operations which call for particular consideration. The first of these is the way in which the title-pages were printed in two colors of ink. Several faint but distinct red ink overprintings of type meant to appear only in black, one of them a full line, indicate first that two machinings took place, one for the letterpress in black ink, the other for the letterpress in red ink; and second that the letterpress appearing in red ink was printed by means of raised type in a forme which also contained the letterpress meant to appear only in black. This method is the normal one as described by Moxon[36] except that the pressman neglected to cut a specially designed frisket to prevent the red ink printing from type which had previously printed in black.

The second operation worth attention is that involving the stop-press substitution of Moseley's imprint for Herringman's and the attendant error of the pressman which produced a hybrid imprint. The pressman, having wrought off his sheets black, an operation in which he stopped his press only once for the substitution of the Moseley imprint in place of that of Herringman, returned the forme to the stone where the compositor removed certain rows of quads in favor of the raised type to appear in red ink. With the second or red ink machining underway for the Herringman title-pages of Parts Three and Four, the pressman needed to watch only for the last sheet bearing the Herringman black ink imprint in order once again to stop his press for the substitution of Moseley's name in place of Herringman's. Evidently in the instance of the Yale University Library Part Three title-page, the pressman either failed to stop the press immediately after he had wrought off the last sheet bearing the Herringman imprint in black or failed to notice a stray sheet bearing the Moseley imprint in black mixed among the sheets carrying the Herringman black ink imprint. Whatever the circumstance, the error was made and later rectified with the following results: there exist two correct states of the title-page, (1) Herringman's name in red ink in the correct black ink imprint and (2) Moseley's name in red ink in the correct black ink imprint. The third state is the hybrid with Herringman's name in red ink in the black ink portion of Moseley's imprint.[37]

4.

The final edition of Parthenissa is the folio reprint of all six parts printed by Thomas Newcomb, Sr., and published by

Herringman in 1676. The volume presents one minor bibliographical problem which is clear cut and readily lends itself to a definite solution.

The text seems to indicate clearly that two compositors set the copy; the problem rests with determining the exact division of their labors. Evidence pointing to the work of two compositors is of several sorts. First to be noted is an abrupt break in pagination between Parts Three and Four, a skip in numbering from page 403 to 485, accompanied by a similar interruption in signatures, a skip from 3F2 to 3Q1, evidently the result of a very inaccurate job of casting off copy. Second, from gathering 3Q to the end of the volume there occur two pairs of running-titles—each compositor used two skeletons in imposing the folio gatherings in fours—entirely different from those appearing in the first portion of the volume. Third, the type-page for all full pages from 3Q to the end contains consistently 53 lines (24 1/2 cm.) while that of all full pages in the first part of the volume contains consistently only 52 (24 cm.).

From this positive evidence we may conclude that the text was divided between two compositors, A and B: compositor A beginning with signature B, the first gathering with normal running titles; compositor B starting arbitrarily with 3Q, the first leaves of Part Four. The variations in running-titles and in the number of lines to the full page indicate that the type set by each compositor was printed in different sets of formes. If the compositors were setting simultaneously, as probably they were, two presses would have been used of necessity.

In addition to these general conclusions, however, there is this one point to be noted. One stretch of 23 pages, extending from page 377 (3C1) to page 403 (3F2) in the first portion of the volume set by compositor A, exhibits all the characteristics of the work of compositor B and of the formes which his pressman was using. Each full page has 53 lines of letterpress, and all the running-titles, starting with page 377 or signature 3C1 and running to page 403 or signature 3F2 are different from all

those of the rest of the first portion of the volume and identical with all those of the second portion of the volume.

Thus the general conclusions for the setting up and printing of the folio text are these: the text was divided between two compositors, A and B, working simultaneously, each with a different press. Compositor A set the type for pages 1 (B1) to 376 (3B4v). Compositor B set the type for pages 485-808 (3Q1-5K3v), the whole of the text of Parts Four, Five, and Six. Since the composing assignment of compositor B was 53 pages shorter than that of compositor A, compositor B very likely finished his task before his fellow workman, and then undertook to help compositor A complete his stint by setting up the text of the last 23 pages (pages 377-403), 3C1-3F2 of Part Three. It is unreasonable to conclude that compositor B completed Part Three before starting on the text of the last three parts; if he had, the abrupt break in pagination and signatures would not have occurred.

5.

The results of this study are summarized in the following revised listing of the various editions, states of editions, and variant title-page imprints of the romance Parthenissa.

• Parthenissa: (Part One, Books 1-8; Part Two, Books 1-8) Quarto [Printed by Peter de Pienne, Waterford] 'Printed in the yeare 1651.' [First issue of first edition] Location: University of Texas Library.
• Parthenissa: (Part One, Books 1-6) Quarto For Richard Lownes, London, 1654. [Second edition] Locations: British Museum; Huntington.
• Parthenissa: 'In Four Parts The First Part' (Part One, Books 1-4) Quarto [Printed by Peter de Pienne, Waterford] For Henry Herringman, London, 1655. [Second issue of first edition] Locations: Bodleian, British Museum, University Library Cambridge, Dulwich College; Huntington, Library Company of Philadelphia, Yale.

  ---

  136

  
  
  
  Variant title-page imprint: For Humphrey Moseley, London, 1655. Location: Huntington.
• Parthenissa: 'In Four Parts The Second Part' (Part One, Books 5-8) Quarto [Printed by Peter de Pienne, Waterford] For Henry Herringman, London, 1655. [Second issue of first edition] Locations: Same as for 'The First Part,' except for Dulwich. Variant title-page imprint: For Humphrey Moseley, London, 1655. Location: Huntington.
• Parthenissa: 'In Four Parts The Third Part' (Part Two, Books, 1-4) Quarto [Printed by Peter de Pienne, Waterford] For Henry Herringman . . . at the Anchor in the Lower Walk of the New Exchange, 1655. [Second issue of first edition] Locations: Bodleian, British Museum, University Library Cambridge; Huntington, Smith College Library. Variant title-page imprints: (*) For Humphrey Moseley . . . at his Shop at the sign of the Prince's Arms in St. Pauls Churchyard, 1655. (†) 'For Henry Herringman . . . at his Shop at the sign of the Prince's Arms in St. Pauls Church-yard, 1655.' Locations: (*) Huntington; (†) Yale.
• Parthenissa: 'In Four Parts The Fourth Part' (Part Two, Books 5-8) Quarto [Printed by Peter de Pienne, Waterford] For Henry Herringman, 1655. [Second issue of first edition] Locations: Bodleian, British Museum, University Library Cambridge; Huntington, Smith College Library, Yale.
• Parthenissa: 'The Fifth Part' (Part Three, Books 1-4) Quarto Printed by T[homas] R[atcliffe] and E[dward] M[ottershead] for Henry Herringman, London, 1656. [First edition] Locations: Balliol College, Oxford; Bodleian; British Museum; University Library Cambridge; Huntington; Smith College Library.
• Parthenissa: 'The Sixth Part' (Part Three, Books 5-8) Quarto For Henry Herringman, London, 1669. [First edition] Locations: Bodleian, University Library Cambridge; Huntington.

---

137




• Parthenissa: 'The Six Volumes Compleat' (Parts One, Two, and Three) Folio Printed by T[homas] N[ewcomb] for Henry Herringman, London, 1676. [Third edition of Part One, Books 1-6; second edition of Part One, Books 7-8, Part Two, and Part Three] Locations: British Museum; University Library Cambridge; Dyce Collection, Victoria and Albert; Haigh Hall, Wigan: Amherst, Clark Library, Library of Congress, Massachusetts Historical Society, Newberry Library, Peabody Institute, and the following university libraries: Chicago, Cincinnati, Columbia, Harvard, Illinois, Michigan, Oregon, Pennsylvania, Princeton, Virginia, and Yale.

Notes