University of Virginia Library

Signatures and Dashes in Novels Printed by T. C. Newby in the Eighteen-Forties
G. D. Hargreaves [*]

Thomas Cautley Newby is best known as the publisher of the first edition of Wuthering Heights (London, 1847), but Emily Brontë's masterpiece was just one of about 130 novels published by Newby in the eighteen-forties.[1] Their usual format is a royal duodecimo,[2] and they are mostly in


Page 254
the standard three volumes, comprising an output rivalled, even in this great decade of London fiction publishing, only by Henry Colburn and Richard Bentley.[3] Many of the volumes, moreover, like volume II of Wuthering Heights itself,[4] have the imprint of Newby not only as publisher but as printer.

These duodecimo volumes with the printer's imprint of Newby tend to show recurrent production features of considerable peculiarity. The most striking of these is the employment of a signature with the figure 3 (e.g. 'B3') on the second of the twelve leaves in the gathering. It is usually accompanied in the gathering by two normal signatures, one on the first leaf ('B'), the other on the fifth ('B5'). The aberrational usage on the second leaf has been recorded in descriptions of Wuthering Heights,[5] but no explanation of the error appears to have been advanced, and it in fact occurs in over 50% of fiction volumes with the printer's imprint of Newby in the period 1846-1848, and in nearly 80% of such volumes in 1842-1843.

Several standard systems for signing duodecimos can be found in the nineteenth-century printers' manuals. Savage, whose generally up-to-date manual of 1841 occasionally reflects eighteenth-century practice, repeats the Moxon-Smith-Luckombe system of placing a signature on every leaf in the first half of the gathering ('A'-'A6').[6] As early as 1808, on the other hand, Stower's scheme for 'twelves' with cutting calls for signatures on the first, second, and fifth leaves only ('A A2 A5' in Stower's diagram, 'B B2 B3' in his text);[7] the rationale here is that the isolated 'A5' or 'B3' especially signals the first leaf of the offcut. Johnson's manual of 1824 shows the same rationale,


Page 255
but here the first leaf of the offcut is signed with the figure 3 in both diagram ('a3') and text ('B3').[8] This use of the figure 3 on the fifth leaf of the gathering has itself seemed aberrational to twentieth-century bibliographers, who naturally equate a signature 'B3' with the leaf B3,[9] but nineteenth-century printers doubtless contented themselves with the logic of signing with the figure 3 the third leaf bearing a signature. Another system, which I have found only in American authorities (e.g. MacKellar), uses only two signatures—one on the first leaf of the gathering and the second (with figure 2) on the first leaf of the offcut ('B B2').[10]

In actual practice, some early nineteenth-century British duodecimo novels, as we might expect, follow the 'B'-'B6' system of the eighteenth century, but all four systems described above appear in British novels during the first quarter of the nineteenth century; although 'B'-'B6' can still be found in the eighteen-forties and 'B B2' also appears then, the second quarter of the century tends to be dominated by 'B B2 B5' and 'B B2 B3', with the latter seeming to gain ground in the eighteen-forties and perhaps becoming the prevailing system after 1850.[11] A novel printed in more than one shop was liable to show more than one system, according to shop usage, as can be seen from the first edition in book form of Oliver Twist (3 vols. 1838), volume III of which has the 'B B2 B5' system in the portions printed by Samuel Bentley, but 'B B2 B3' in the portions printed in Charles Whiting's shop.[12]

The 'B B3 B5' aberration practised in T. C. Newby's shop looks like a conflation or confusion of these normal 'B B2 B5' and 'B B2 B3' alternatives, and a direct source of the confusion is actually to be found in the manuals of Hansard (1825) and Timperley (1838). The fact that Stower (1808), as we


Page 256
have seen, has 'B B2 B3' in his text and 'A A2 A5' in his diagram possibly did not help his successors, but at least these were cogent alternatives. Hansard, on the other hand, has 'B B2 B5' in his text, but his diagram shows 'A A3 A5'.[13] Timperley, for his part, has 'B B2 B3' in his text, but gives exactly the same diagram as Hansard, even though it illustrates an alternative system and does so with an error.[14] When the textbooks were capable of such things, and when at least two cogent systems were in regular current use, it is hardly surprising that some practitioners were led astray.

That personnel setting novels in Newby's composing room were often ill-equipped to combat such misleading and confusing circumstances is suggested by several further peculiarities evident in the signatures which they set. The most consistent of these peculiarities is the placement of signatures decidedly to the left of centre in the signature line. This seems to be especially the case with the signatures on the second and fifth leaves, the one on the first leaf often being more centrally placed (probably to avoid close proximity to the volume number). Instruction on signature placement is rather scarce in the printers' manuals, but such as can be found more or less favours the right. Moxon (p. 210) prescribes "about the middle of the Line, or rather a small matter nearer the end than the middle is, (because when the Sheets are wrought off and gather'd, they Collation something quicker . . .", while Johnson (II, 134) states: "In our opinion the signature is much better both in appearance, and for collating, when placed within about four m's from the end of the line, than in the centre, according to the old custom." Timperley (p. 17) goes even further and specifies "the right hand corner of the page." In actual British practice in the second quarter of the nineteenth century, signatures do seem to be placed most frequently more or less to the right of centre, with occasional central placement, and any tendency more or less to the left, as seen in these volumes printed by Newby, would seem to be very unusual at this and indeed most other periods.[15] Apart from this peculiarity of signature placement, random mis-signings (in addition to the basic 'B3' for 'B2' error) occur in Newby printing, while his signatures are frequently omitted and occasionally even inverted ('B3'), reversed ('3B'), or placed on a verso.[16]


Page 257

Another production feature characteristic of novels printed by Newby, especially in the period 1845-1847, is the intermittent use of three hyphens ('---') or three full points ('. . .') as a substitute for the em dash ('—'). The reason for this is not altogether clear. Some novels printed by Newby seem to be particularly rich in the use of the dash (or its Newby-style substitutes), and it is possible that a shortage of em dashes proper may have arisen from a tendency on the part of Newby's compositors to set dashes strictly according to copy when other shops would have substituted more formal punctuation. Unfortunately, there appears to be no surviving printer's copy for any Newby novel, but comparison between the printer's copy and the first edition of Jane Eyre (printed by Stewart and Murray, for Smith, Elder, 3 vols. 1847) shows the considerable extent to which an author's dashes could be suppressed by the printer in favour of more formal punctuation.[17]

Not surprisingly, Newby's own press was unable to cope with all his publishing commitments, and the imprints of other printers appear in some of the novels which he published. But many of the fiction volumes with his publisher's imprint have no printer's imprint at all,[18] and the combined appearance in some of these of the peculiar features described above would seem to constitute a useful body of evidence that he was also their printer.[19] Thus while only the second of the two volumes of Wuthering Heights (1847) carries the name of Newby as printer, it seems almost certain that volume I, with the 'B B3 B5' signing system, an inverted signature, many signatures placed on the left, and the occasional use of '. . .' for '—', is also a product of Newby's printing shop.[20] A similar case is Charles Lister's The College Chums (2 vols. 1845), in which only volume II has a printer's imprint (Newby's)


Page 258
but both volumes use 'B B3 B5', with many signatures on the left and occasional '---' for '—'. Sybil Lennard (3 vols. 1846), attributed to Mrs. Grey,[21] has no printer's imprint in volume II (volumes I and III have Newby's), but all volumes use 'B B3 B5', with signatures generally on the left, while volume I has occasional '---', and both '---' and '. . .' appear frequently in volumes II and III. Conversely, it seems certain that Newby was not himself involved in the printing of Anne Brontë's The Tenant of Wildfell Hall (3 vols. 1848), which throughout uses 'B B2 B3', with all signatures on the right and all the many dashes normal,[22] and although volumes I and II are without a printer's imprint, there is no reason to suppose that they were not printed in the same shop as Volume III, which has the printer's imprint of J. Billing, Woking.

It may be thought unnecessary in these cases to suppose that there might be any question of an imprint in just one or two volumes not speaking in itself for an entire set, but two further novels published by Newby readily illustrate the potential problem. Although the final volume of Anthony Trollope's first novel, The Macdermots of Ballycloran (3 vols. 1847), has the printer's imprint of Newby and 'B B3 B5', with signatures central (first leaf) or on the left (second and fifth leaves) and a bout of '---' and '. . .' for '—' (pp. 128-131), it seems unlikely that Newby printed volumes I and II, which have no printer's imprint but use the 'B B2' system, with signatures central or on the right and normal dashes. Again, while the third volume of Eliza Lynn Linton's Azeth the Egyptian (3 vols. 1847) has the printer's imprint of W. N. Judd and 'B B2 B3', with signatures all well to the right and normal dashes,[23] volumes I and II, which are without a printer's imprint but have 'B B3 B5', with signatures mostly on the left, several unsigned offcuts, and frequent '. . .' or '---' for '—', were surely printed not by Judd but by Newby himself.



Most of the novels in question have been examined in the Bodleian and/or the National Library of Scotland, and I should like to thank the reader-services staff in these libraries. Reader-services staff at St. Andrews University Library have also been very helpful. The greatest acknowledgement is due to my wife, without whose extensive research assistance this note could not have been compiled.


Figure based on a survey of Andrew Block, The English Novel 1740-1850: A Catalogue, new and rev. ed. (1961). Newby appears to have commenced publishing in 1842 in association with T. & W. Boone and to have become established in his own right in 1843. His output was by no means confined to fiction. (See further The Publishers' Circular and The English Catalogue of Books.)


Leaf size approx. 7 x 4¾ ins. in uncut copies. Gatherings are of twelve leaves, and unopened copies show that duodecimo with cutting was employed, the four central leaves of the gathering comprising the offcut. Contemporary trade sources (such as The Publishers' Circular and The English Catalogue) tend to describe these novels as "post 8vo."—true post octavos are in fact rather similar in leaf size to royal duodecimos but are of course differently produced.


Figures emerge from Block of about 130 for Colburn and about 110 for Bentley. Other noted publishers of fiction are well down on these figures for the period (Longman 55, Chapman & Hall 40, Saunders & Otley 35, Smith, Elder 25).


Wuthering Heights appeared as volumes I and II of a hybrid three-decker with Anne Brontë's Agnes Grey as volume III, which also has the printer's imprint of Newby. Volume I has no printer's imprint—see below.


T. J. Wise, A Bibliography of the Writings in Prose and Verse of the Members of the Brontë Family (1917; facsimile impression, Dawsons, 1965), p. 98; Albert Smith, in the Clarendon Edition of Wuthering Heights, ed. Hilda Marsden and Ian Jack (1976), p. xxxiv.


Joseph Moxon, Mechanick Exercises on the Whole Art of Printing (1683-4), ed. Herbert Davis & Harry Carter, 2nd ed. (1962), pp. 211, 225(diagram); John Smith, The Printer's Grammar (1755; facsimile impression, Gregg Press, 1965), diagrams on pp. 236-237 (but the text, on p. 208, calls for only five signatures); Philip Luckombe, The History and Art of Printing (1771; facsimile impression, Gregg Press, 1965), pp. 382, 413(diagram); William Savage, A Dictionary of the Art of Printing (1841; facsimile impression, Gregg Press, 1966), p. 340(diagram)—Savage's diagram takes 'B' signatures for the example, as does the present note when not quoting manuals that use 'A' (these Newby novels do not normally have any 'A' signatures; typically they commence with an unsigned singleton—the title-leaf—followed by gathering B12).


Caleb Stower, The Printer's Grammar (1808; facsimile impression, Gregg Press 1965), pp. 163, 174(diagram). For twelves without cutting Stower gives signatures on the first, second, and third leaves ('B B2 B3')—p. 175(diagram).


John Johnson, Typographia (1824; facsimile impression, Gregg Press, 1966), II, 135, 158(diagram *14).


Wise, A Bibliography . . . of the Brontë Family, p. 109, notes under the first edition of Anne Brontë's The Tenant of Wildfell Hall (3 vols. 1848): "In each volume the signature to the fifth leaf of each sheet is usually misnumbered 3" (see further on The Tenant below). R. B. McKerrow called it "a curious and very misleading method," in An Introduction to Bibliography for Literary Students (1927), p. 79. Mrs. C. G. F. Gore's The Man of Fortune (3 vols. [1842], printed by Charles Reynell for Colburn) has not merely the third signature (e.g. 'B3') on the fifth leaf but the second signature (e.g. 'B2') consistently on the third leaf.


Thomas MacKellar, The American Printer, 13th ed. (1882), p. 165(diagram); American Dictionary of Printing and Bookmaking (1894; facsimile impression, Detroit, Gale, 1967), pp. 279(diagram), 509—this actually refers to half-sheet imposition for gathering in sixes, but the principle is the same.


Based chiefly on a survey of the substantial holdings of nineteenth-century novels in St. Andrews University Library. 'B'-'B6' occurs in Theodore Hook's Fathers and Sons (3 vols. 1842, printed by Gilbert & Rivington for Colburn)—this also appears to have residual press-figures, the latest I have seen and another relatively out-of-date feature of Savage's 1841 manual (p. 814); see also note 19 below. 'B B2' was used by James Ballantyne in Edinburgh for e.g. Scott's Waverley (3 vols. 1814), Guy Mannering (3 vols. 1815) and The Antiquary (3 vols. 1816), as was noticed by R. W. Chapman in The Library, 4th ser., 4 (1924), 180 note; in the eighteen-forties it occurs in Trollope's The Macdermots of Ballycloran (see below).


See Kathleen Tillotson, in the Clarendon Dickens edition of Oliver Twist, ed. Kathleen Tillotson (1966), p. xlix.


T. C. Hansard, Typographia (1825; facsimile impression, Gregg Press, 1966), pp. 498, 508(diagram).


C. H. Timperley, The Printers' Manual (1838; facsimile impression, Gregg Press, 1965), pp. 18, 26(diagram).


A rightward tendency appears to be usual up to c. 1700; from c. 1725-c. 1825 a preference for central placement is evident, with a return to a rightward tendency c. 1825-c. 1900. In modern British practice, however, signatures (if used at all) tend to be placed on the extreme left.


Errors of this kind may not, of course, persist through an entire edition, though five copies of C. F. Henningsen's Sixty Years Hence (3 vols. 1847), including what appear to be two copies of a second issue—calling itself "second edition"—1847 and one of a third issue 1848, do show some such anomalies in common (my thanks for information to Ms. Jessica S. Owaroff of the Houghton Library, Harvard University; Mr. Thomas D. Burney of the Library of Congress; and Mr. Douglas Matthews of the London Library); but see note 20 below. Some hints can be found of the distraction these errors must have caused in warehouse and bindery: the National Library of Scotland copy of volume I of Sixty Years Hence (second issue, original cloth) has no fewer than ten of the offcuts inserted in the wrong gatherings.


See the textual footnotes in the Clarendon Edition of Jane Eyre, ed. Jane Jack and Margaret Smith (1969), esp. p. 408.


The strict legal sanctions against this introduced in an Act of 1799 had been considerably reduced by Acts of 1811 and 1839—see W. B. Todd, "London Printers' Imprints, 1800-1840," The Library, 5th ser., 21 (1966), 46-59.


The combination factor must be stressed: a few examples of '---' for '—' occur in volume III of Lady Charlotte Bury's Family Records (3 vols. 1841, printed by B. Bensley for Saunders and Otley), though the signatures are on the right and normal (except that volume I has 'B'-'B6' but volumes II and III use 'B B2 B5', in spite of the fact that all have the same printer's imprint); again, the anonymous Roland Percie (2 vols. 1846), a novel published by Newby but with the printer's imprint of G. Paul, Bury St. Edmunds, in both volumes, has the 'B B3 B5' system with some signatures placed on the left, though the dashes are normal throughout.


Albert Smith, in the Clarendon Edition of Wuthering Heights, p. xxxiv, indicates that the inverted signature ('B3' on B2) does not occur in all copies. It may be significant that '. . .' for '—' proliferates in a section of volume I where one might particularly expect the copy to have employed heavy use of the dash—the delirious assertions of Catherine in chapter 12 (some modern editions—e.g. Clarendon, World's Classics, Penguin English Library—commendably, but I believe wrongly, retain '. . .' here).


Full name in catalogues given as either Catherine Maria Grey or Elizabeth Caroline Grey.


Like some novels printed by Newby, this first edition of The Tenant may well retain a relatively high proportion of dashes from the lost MS. copy—the printed punctuation, especially in dialogue, is often very informal.


This is not to say that volume III of Azeth is generally well printed—like Wuthering Heights its text shows numerous disfigurements attributable to case and press, among the more striking of which are several instances of '!' for 'l' (e.g. 'credu!ity') and the occurrence of 'Thy no sojourn' for 'Thy sojourn on'. In spite of this, it is possible that, in some respects, shops like Newby's and Judd's, by their very lack of sophistication, may have preserved features of lost MS. copy to a greater extent than more competent establishments.