University of Virginia Library

Thomas D'Urfey's Richmond Heiress (1693): A Bibliographical Study
Raymond A. Biswanger, Jr.

The first edition of Thomas D'Urfey's comedy, The Richmond Heiress: Or, A Woman Once in the Right, presents a number of interesting bibliographical problems. It was first published in 1693, evidently from a manuscript which had served as a prompt copy or, rather more likely, which had been marked to serve as the basis for a transcribed prompt book. An actor's name appears sporadically in place of the character he portrayed; more important, warnings to the actors about to enter the scene have crept into the text in two places.[1] The quarto collates A-I4, pp. [8] 1-32

29-36 41-64 (all pp. in parentheses except for 57-64 in square brackets). The book appears to be very carelessly printed, very likely in considerable haste.

This edition exists in two distinct issues, the second of which, as we shall see later, is a reimpression except for two sheets.

The first issue, distinguished as (i), exists with the outer forme of sheet A, including the title-page, in two variant states. In the earliest pulls of this forme (iτ) the title-page reads as follows:

Comparatively early in the printing (on the evidence of the few copies preserved of this early state) the press was stopped. On the title-page the misprint 'ROAYL' was corrected to 'ROYAL', and the line attributing the authorship to D'Urfey was revised to read

This produces state (i*), performed by press-correction with the remainder of the type of the title left undisturbed.

The alteration of the line containing the author's name has an interesting background. In 1691 D'Urfey had been scathingly attacked in the epistle dedicatory of the anonymous pamphlet Wit for Money, in which he was referred to as, "a certain Poet, who before the Poll Acts, used to write himself T. D. Gentleman. . . ." The original form of the Heiress title-page seems to have been the work of the printer, therefore, and the change made by the author himself. This view is substantiated by important textual changes made elsewhere in the outer forme of this sheet at the same time as the revision of the title-page. On sig. A2v 'The Scene Richmond-Hill.' was added beneath the dramatis personae. On A4v a necessary correction was made in the text in stanza 3 of 'SHINKEN's Song to the Harp' by the alteration of 'Highfoot' to 'light-foot'. On the same page in 'SONG. In the Last ACT' the word 'Philosophers' was altered to 'Philosophey' in the last line of stanza 3; in the third line of stanza 6 'A Health' became 'Success'; and simultaneously an entire new stanza, a seventh, was added to this song. These are clearly authorial revisions.

During the course of impression in this first issue, one small press-correction was made in sheet E, when on sig. E2 in the last line 'I must needs, nay,' was altered to 'I must needs say,'. However, both formes of sheet H were heavily press-corrected on every page. A total of ten variants appear in the inner forme, the most convenient identification being in line 12 of H1v where the stage-direction '[to Doggett.' of the early pulls was altered in the corrected state of the forme to '[to Quickwit.'. Most of

these variants represent correction of egregious compositorial misreadings and lapses, but several represent literary revision. In the outer forme of H appear twelve similar corrections. In addition to these the warning to the actors, '[Sir Charles.T.Romance, Shink.Guiac.Constable.' at line 5 of H1 in the uncorrected state is eliminated in the corrected. The two stage directions on H1 (lines 24-25 and 29-31) are reset, the first being made less crowded, and the second indented from the left margin, this last producing consequential changes in the lineation. In all the copies observed, both formes of sheet H were either in their uncorrected or else in their corrected states; in no case was an uncorrected forme backed by a corrected, or vice versa.

It is an equally curious fact that owing to the distribution of the sheets in the binding process, every observed copy of issue (iτ) with uncorrected sheet A (containing the 'ROAYL' title) is bound with uncorrected sheets E and H. Moreover, every observed copy of issue (i*) with corrected sheet A ('ROYAL' title) was bound either with uncorrected sheet E but corrected sheet H,[2] or else with corrected sheet E and uncorrected sheet H.[3] No observed copy of (i*) contained both E and H uncorrected or both corrected.

Before passing on to the second issue (reimpression) and its variants, I should say something of the printing of issue (i), since certain facts in relation to this have a bearing on the interpretation of the peculiarities of issue (ii). The evidence is, unfortunately, somewhat contradictory and at first sight not susceptible to an exact bibliographical solution.

For example, if we took only the evidence of spelling tests to determine compositors, we should divide the book into sheets B-E set by compositor X, and sheets F-H set by compositor Y. What evidence there is would assign sheet A to workman X. As for sheet I, it is quite anomalous since it contains some characteristics of both X and Y. A marked compositorial break seems to occur beginning with sheet F.

                                     

Compositor	X				Y			?
Sheets	B	C	D	E	F	G	H	I
I'll	15	15	9	22	12	5	17	18
I'le (or Ile)					3	7	3
ile					2
intrigue (and forms)	6	3	5	1				2
intreague (and forms)					3	4	7
'Gad	7	1	3	6
Gad					3	12	11	7
Quickwit	6	2	6	3	1	2	1	9
Quickwitt					2	2	2
Marmalet	1	1	5	3	1			1
Marmalett						5	1
Hotspur	2	4		1	6		1
Hotspurr					2		1	2
Shinken	2	1	3	1
Shinkin	4				3	3	6	3
Roman scene numbers	1	2	1	1	1
Arabic scene numbers					1	3	4

This table seems to point to only one compositor of sheets B-E, a workman of regular characteristics. Correspondingly, at least for sheets F-H, another compositor seems to be present, one of mixed characteristics but nevertheless in the majority of cases differentiating himself from the first workman.

On the basis of four appearances of the spelling 'intrigue' (and forms) in the dramatis personae, there are grounds for assigning sheet A to compositor X. Sheet I is troublesome. The compositor shares with X the invariable spelling 'I'll', the invariable spelling 'intrigue' (once on I2v 'intriegue'), and the spelling of 'Quickwit' and 'Marmalet' with single final consonant. On the other hand, he doubles the last consonant in 'Hotspurr' like Y; also like Y he does not place an apostrophe before 'Gad' and he employs the invariable spelling 'Shinkin'. A breakdown of the occurrences of these variants by pages does not encourage a hypothesis that X and Y alternated. Although X could have set I1r, and Y could have set I1v-2v, on I3-4r there is the clash between Y's 'Gad' without apostrophe, one to each page, and X's spellings of 'Quickwit', 'Marmalet', and 'intrigue'. Possibly another compositor, Z, is present, a workman of mixed characteristics in following copy.

It should also be noticed that the compositor of sheet F, though in most respects associated with the characteristics of compositor Y, especially in such important invariables as 'Shinkin', 'intreague', and 'Gad' without apostrophe, yet is heavy on the 'Hotspur' spelling, once uses (on sig. F1r) a roman scene number, once the 'Marmalet' spelling, and is unique in setting parentheses about stage directions instead of the square brackets invariably used elsewhere in both sections. The odds are that he is Y, but some doubt may exist.

The evidence for a sharp break in compositorial characteristics occurring between sheets E and F strongly suggests the hypothesis that this book, like so many Restoration play quartos, was divided approximately in half between two compositors and presses and the two sections simultaneously typeset and printed. Moreover, this evidence is buttressed by the significant fact, commented on below, that in the reimpression which constitutes issue (ii) both sheets B and F (the start of the two sections of text) show no signs of reimpression and are definitely sheets from the same run as issue (i). This further serves to differentiate the two sections of the book, and may be taken as confirming the general hypothesis for simultaneous printing in two major sections. Yet it is by no means certain that only two compositors and presses were involved. Sheet I, for example, with its unique headline square brackets and its mixed spelling characteristics seems to be associated with a different compositor and press, brought in to help finish the book as quickly as possible. In the second section the two preceding sheets G and H definitely are the product of only one workman, on the evidence of the spelling but more important on the evidence of their similar type-page measurement of 48 ll. 194-5 (205-6) x 122 mm. But sheet F, already mentioned as slightly irregular in spelling, not only has smaller page numbers in the headlines than those in sheets G and H but also was set to fit a quite different skeleton-forme, its average type-page measurement being 46 ll. 188(198) x 126. Moreover, 20 lines of its type measures about 82 mm., whereas the type measurement for sheets G and H is about 81 mm. Thus sheets G-H would seem to have been set by a different compositor from that of F (which on the whole is difficult to believe), or else there was a delay in the printing after the composition and machining of sheet F.

In the first section there seems to be a similar break. Sheets B and C have the same small numbers in the headline, approximately the same type-page measurements, and seem to have been set with the same printer's measure of 119-120 mm. and in the

same font of type. Yet sheets D and E change to larger figures in the headlines, and are set to fill a different skeleton-forme, since their measure changes to 126 mm.; moreover, though sheet E has apparently the same font of type as B-C, about 81 mm., that in sheet D is an 82 mm. type. Since the spelling tests make it almost impossible to conjecture more than one compositor, clearly this is an anomaly. Possibly a delay between C and D can be posited here, as could be conjectured between F and G; but if so, it is a little more difficult to explain why all the observed copies of issue (ii) contain sheet C re-impressed.

Issue (ii), with the exception of sheets B and F, was printed from the standing type of issue (i) re-imposed and in parts revised. The title-page now reads:

This is the same form as in issue (iτ), and represents the standing type of this issue, although there have been a few typographical alterations. The letter 'R' in 'Richmond' has been changed from one with a straight tail to a sort with a curved tail. The lower rule, broken in (i), has been replaced and the two rules moved nearer each other. In (i) the rules are 25 mm. apart, but in (ii) only 18 mm. The roman 'l' in 'Will's' of issue (i) has been replaced with an italic type. On A4v the spelling Philosophey in the third stanza of the song in the last act has been altered to Philosophy. In the other sheets changes in the alignment of the page numbers in the headline in relation to the text, or alterations in the font used to set the paginal headlines, demonstrates that the type-pages have been stored and reimposed for a second impression. There is no question, however, that sheets B and F in both issues come from a single impression, since there is no alteration in the headlines or in the text. Moreover, the outline of a hair on the face of the type across lines 16-18 of sig. B1v is present in both issues, presumably an impossibility if the type-page had been rinsed and stored between impressions.

Before the impression a few, mostly minor, corrections were made in the standing type. For instance, on C2v, line 25, 'hear you are' in (i) was altered properly to 'here you are' in (ii). That these were the sole responsibility of the printing house is shown by the miscorrection that followed on C3r, line 7, of 'She hear!' to 'She here!' through a misunderstanding of the context. Some few of the differences result from type dropping out, as the disappearance of the u between impressions in the speech-heading on G2v, line 20, from 'Quick.' in (i) to 'Qick.' in (ii). Viewed as a whole, the alterations made in standing type do not suggest authorial intervention. Of the text, such variants appear only on C2v, C3, G2v, G3, H2, H3v, I4v. Grouped thus as formes, it is likely that the slight changes were made just before printing the second impression.

More important changes appear, however, in the second impression, since a certain amount of resetting is found. Thus the whole of sigs. D1v and D2 appear in a new typesetting, as well as the whole of G4v and the last 15 lines of G1. The variant settings may be identified as follows: D1v, line 8, 'four' with a comma in (i) but no comma in (ii); D2, line 26, 'surprized' in (i) but 'surpriz'd' in (ii); G1, line 34, 'disturb' in (i) but 'desturb' in (ii); G4v, line 1, ends with 'Heiress' in (i) but 'Heiress agen' in all but one copy of (ii). The text in these reset portions shows normal compositorial corruption and there is no evidence that any variant has authority.

Within sheets D and G these reset portions are confined to a forme from each sheet. Although it may be that they resulted from accidental pieing in handling the type, another suggestion is possible and will be offered later.

The second issue exists in several variant states. Among the copies examined, Huntington (123058) is unique in possessing an added leaf (bound between A3 and A4) which is paged 61 and 62 and contains a song "The Country Gentlemen" on recto and verso.[4] The pagination is enclosed in square brackets, similar to the practise in sheet I of the text. This addition is accompanied in the Huntington copy by two cancellations made to include revisions to the text. The Gentleman's Journal for November, 1693, noticed: "Mr. Durfey's Richmond Heiress has been Revis'd, and Acted several times, with Alterations and Amendments." The revisions on cancel leaves G3 and I1 would appear to reflect these later changes, at least in part.

In IV.iii on G3r the changes take away comment on Tom Romance's intrigue with Mrs. Stockjobb in order to emphasize the romance with Sophronia which figures so prominently at the end of the play. Various additional disparaging remarks about Tom's father seem to heighten Tom's role as a profligate and so justify the more his final undoing later. In order to insert this new material, G3r was reset in smaller type, with 55 lines to the page as against 48 in the cancellandum, and some of the revision was carried over onto the verso, which is set in type similar to that of the original leaf. Parentheses are used about the headline page numbers in the cancellans.

On sig. I1 the action of V.iv had originally continued with a farcical episode involving Numps. In cancellans I1 this action is excised completely and, instead, Scene Ultima begins at the head of the cancellans leaf recto with an interview between Sophronia and Tom Romance (lacking in the original version) before the entrance of Sir Charles, Guiacum, and Hotspur. Originally, their entrance, with that of Romance, had begun the first version of the final scene on cancellandum I1v. Parentheses are used about the headline pagination of the cancellans.

From the similarity of paper and the fact that the chain-lines match up, it seems reasonably certain that the added song bound in the Huntington copy's preliminaries and cancellans leaf I1 were printed together by half-sheet imposition and then cut apart for insertion. On the other hand, the paper of cancellans leaf G3 seems to differ slightly, and it may be that this represents a separate printing.[5]

The Wrenn copy at the University of Texas also contains variants not found elsewhere. This copy is peculiar in that the title-gathering, sheet A, in the second state of the first impression (i *) is bound up with sheets B-I of issue (ii). Presumably

there were a few extra copies of this sheet and it was economically held over to assist in making up the second issue. Sheet G contains on sig. G1 a pagination variant, the leaf being numbered 37 and 38 instead of the 41-42 found in all other copies of both impressions. Apparently when the formes were being readied for the second impression the printer noticed the gap between page 36 on sig. F4v and 41 on G1r and essayed to bridge it to indicate to the purchaser that no material was missing. Since there are no other variants on this leaf it is impossible to demonstrate whether this change was in fact made late in the run of the first forme through the press, or early as I have assumed. Whichever it was, there is some interest in the consequential change made in the opposite forme to coincide (on this slight evidence) only with the same sheets of the first forme printed with the variant pagination.

Finally, there is still another variant state in the copy of issue (ii) preserved in the Ohio State University library. This is found in a rearrangement of the type in the first three lines of sig. G4v, one of the reset pages in issue (ii). In the impression represented by issue (i) these lines had read "time to tell ye, that if you follow me quickly, you may recover the Heiress / agen. [Speaks as out of breath. / Sir Char. Hah, — what say'st thou?" In all of the regular copies of issue (ii) in the resetting these three lines were condensed to two, reading, "time to tell ye, that if you follow me quickly, you may recover the Heiress agen. / Sir Char. Hah, — what say'st thou? [Speak [sic] as out of Breath." In the Ohio State copy, however, although the type is that of issue (ii) the lines are arranged as three and the stage-direction reads "Speaks" as in issue (i). The position of the direction in issue (ii) is clearly wrong, since the first speaker (Cunnington) has just said "I have only now Breath and / time to tell ye . . ." and therefore the stage-direction must apply to him and not to Sir Charles. It is obvious, therefore, that the Ohio State copy represents the first state of the type on G4v in the second impression, which reprinted that of the first impression, and that the type for these lines was not rearranged for revision (in the state represented by other copies of the second issue) but instead for some purely mechanical reason. It is most plausible to guess that the lines were reduced from three to two the better to accommodate the type-page in the opening of the skeletonforme, and in the process it was necessary to place the direction after a different line even though in error.

Variants in Issue (ii) [all copies examined]:

• 1) CSmH (123058): added leaf with new song; G3 and I1 cancellans leaves.
• 2) TxU (Wrenn): G1 paged 37-38 (earliest state?). Sheet A(i *) bound with sheets B-I of issue (ii).
• 3) OU: G4v, earliest and most correct state of first three lines on G4v.

Any use of standing-type, especially in connection with partial resetting, calls for explanation; and for The Richmond Heiress this is the more necessary bibliographically because of the peculiarities which are apparent in the printing of the first impression and which have a connection with peculiarities, such as the first-impression sheets B and F, in the second impression. The following hypothesis is offered in full consciousness that there still remain a few details of the printing process which seem difficult to pinpoint exactly.

The date of the first performance for the Heiress is not quite certain. Nicoll assigns it to c. February, 1693; Summers and Harbage to c. March; and Day to c. April. The later date seems much to be preferred, on two counts. First, the play was advertised in The Gentleman's Journal for April, 1693 (p. 130). Second, in a letter from Dryden to Walsh dated May 9, 1693, Dryden wrote:

Since D'Urfey's dedication in the quarto (in which he speaks of the play's failure) is dated May 6, it is very likely that the play was first produced in late April or the first few days of May. In spite of Dryden's prognostication that the farce was "kickd off for ever," the Heiress was revived the following November and apparently then with some success. As previously noticed, The Gentleman's Journal for November, 1693 (p. 374), writes that the play "has been Revis'd, and Acted several times, with Alterations and Amendments."[6]

It seems clear from the way in which many Restoration play quartos were hurriedly printed in two or more sections simultaneously that the first appearance in print ordinarily coincided closely with the date of initial performance, and there is some reason to suspect that authors liked to have copies available for the third, or author's, night of the play. Thus Briscoe, it is almost certain, would have started to print the play shortly before its first night. On the evidence detailed above that sheets B and F were simultaneously printed on two presses, we may suppose that they represent sheets run off in anticipation of an edition of normal size for a successful play. Only on this assumption can we explain the fact that enough copies of these sheets were machined to serve for the second impression without reprinting. It seems legitimate to assume, further, on the basis of the evidence that when the catcalls of the first-night reception[7] demonstrated that the play was doomed to failure, Briscoe reduced the size of his edition-sheet beginning with sheets C and G and put out his first issue in a relatively limited quantity.

However, failure on the stage, it would be possible to argue, need not necessarily mean failure with the reading public. Moreover, D'Urfey was a successful dramatist and may have laid almost immediate plans for revisions in the hope of a revival after the turmoil had subsided. Thus it seems that Briscoe decided, or was persuaded, to keep the type standing in anticipation of a possible further demand in the future which would justify the inconvenience in view of the money to be saved if a second edition were called for. Here the resetting of two pages in sheet D and of one page and a portion of another in sheet G assumes a new significance, for the position of this resetting argues for deliberate and mechanical distribution of the original type after the first impression rather than an accident to type-pages in storage. If this is so, sheet D (the third of the first section) and sheet G (the second of the second section) were probably run off very close to each other in point of time and had begun to be distributed before Briscoe decided to keep any of his type standing for a possible later impression.

The date of the second re-impressed issue is uncertain but may be placed either before the November revival as a consequence of a reading demand for the play or else, more probably, in close connection with the revival itself, probably coincidental with it. That Briscoe was not aware of the changed text which marked the revival is shown by the fact that no authorial changes were made in his standing type. Presumably after the later relative success of the play, a few remaining copies were

brought up to date by the cancellation of leaves G3 and I1 to provide some portion of D'Urfey's revisions, as well as the addition of a new song, as represented, among the copies examined or reported for this study only in the unique Huntington exemplum. If this hypothesis is reasonably accurate, it would seem, then, that Briscoe kept his type standing from early May to November.

The main outlines of this reconstruction do not seem to be controversial and they may roughly represent the facts in all probability. The problems raised by the anomalous evidence for the printing of the first impression may now be surveyed with somewhat more speculation. It will be recalled that a spelling test shows little doubt that sheets B-E (and probably A) were set by one compositor, this comprising the first section of the book. In the second section sheet G and H were clearly set by the same workman; but sheet F, although in general showing his characteristics, is not quite so uniform, and sheet I is rather mixed. Within each section bibliographical evidence suggests some lack of continuity in typesetting and also in printing. Thus whereas sheets B and C employ the same font of type and the same printer's measure, and utilize the same font for the pagination in the headline, sheets D and E coincide in a different font for the pagination, and in a different printer's measure (126-127 mm. as against 119-120 mm. for B-C), and differ only in the font for the text, which is slightly larger in sheet D than that employed in sheets B, C, and E. Correspondingly, sheet F in the second section is unique in its treatment of setting off stage-directions by parentheses instead of square brackets, in the small size of its headline pagination, in the longer printer's measure (125-126 mm. versus 122-123 for G-H, and 121-122 for I) and in the slightly larger size of the font used for the text. Sheet I differs in its use of square brackets instead of parentheses about the headline pagination; and though its font seems to be the same as that in sheets G and H, its measure is slightly narrower.

When we put all these facts together we may conjecture that in the original impression sheet F was started somewhat later than sheet B. The break that exists between sheets C and D, and between F and G, suggests that when Briscoe was apprised of the play's failure sheet F was on the press at the same time as either the last of sheet B or the early pulls of the first forme of sheet C. At any rate, the compositor of the first section was far enough in advance of the second-section workman to have completed his setting of sheet C, but the second compositor had not, for some reason, begun to set G. Thus the simultaneous break in the typesetting after C and F may further suggest that for a short time Briscoe suspended work on the play, possibly with a view to cutting his losses at that point by giving up publication. He then, seemingly, changed his mind, and began again by machining sheet C in a reduced edition-sheet (from the type already set) and starting fresh composition with sheets D and G.[8] If this is so, the irregular sheet I is explicable as the work of a third compositor and press, since the first press would be working on sheet A when the second was engaged with sheet H, and to complete the play most expeditiously some extra assistance was required in the second section.

The only irregularity remaining is the shift in sheet E back to the font of type used in sheets B and C instead of the slightly larger font employed for sheet D when setting was resumed. This is mere guesswork, but it may not be unreasonable to

conjecture that since sheet E had to join with the text for sheet F, long since printed, the inadvertent choice of type for sheet D was corrected in order to set more text, as was necessary, in sheet E.[9]

Notes