 | Notes on Recent Work in
Descriptive Bibliography |  |
PAPER
My essay on "The Bibliographical Description of Paper" was pub-
lished in Studies in Bibliography, 24 (1971), 27–67, and
reprinted in Readings
in Descriptive Bibliography, ed. John Bush Jones (1974), pp. 71–115, and in
my Selected Studies in Bibliography (1979), pp. 203–243.
I was pleased that
my approach was approved by the great scholar of paper, Allan Steven-
son, whom I had consulted shortly before the end
of his life, when we were
ars have made the most significant advances in the study of paper from a
bibliographical point of view: David L. Vander Meulen, Paul Needham,
B. J. McMullin, and John Bidwell.
Of their publications, the one with the widest applicability to biblio-
graphical
description, and a fundamental essay for the analysis of paper, is
Vander Meulen's "The Identification of Paper without Watermarks:
The
Example of Pope's Dunciad," in Studies
in Bibliography, 37 (1984), 58–81.
(Three years earlier he had
discussed and illustrated the various tech-
niques taken up here in his pioneering
1981 dissertation, a descriptive
bibliography of Pope's Dunciad; see pp. 47–58 for his general account
and pp. 72–77
for one outstanding example.) Most previous bibliographi-
cal work on paper
(including mine) had emphasized watermarks and the
chainline spaces close to them
and had either stated or implied that all
the other chain spaces were equal to each
other. But Vander Meulen
points out that the spaces
between chains in handmade paper vary within
a given mould (and from one mould to
another) and that the sequence
of the distances between chainlines can identify the
sheets produced in
a single mould. Although for various reasons a given space may
not be
consistent in all those sheets, the variation is slight and the overall
pattern
is not affected.
In describing books printed on handmade paper, therefore, it is not
sufficient
simply to say, for example, "chainlines 23 mm. apart." Instead,
the whole sequence
across a sheet ought to be recorded, and Vander
Meulen has
supplied a system that bibliographers should adopt. He uses
vertical lines to
separate chain-space measurements (extending the prac-
tice Stevenson follows near watermarks), with three further
conventions:
parentheses indicate an estimated or incomplete measurement (at a
gut-
ter or when a sheet was cut between chainlines); ellipsis dots signify
an
incomplete sequence (for a partial sheet); and the percentage sign stands
for
a deckle edge (in parentheses when the deckle is inferred). A record
(with
measurements to the nearest half-millimeter, taken at the middle of
the sheet) might
look like this:
(9) | 19 | 29.5 | 27.5 | 31 | (29) | 29.5 | 30 | 28.5 | 20 | (13)
(One desirable modification suggested by John A. Lane is to
use a verti-
cal wavy line instead of the percentage sign: see "Arthur Nicholls and
His Greek Type for the King's Printing House," The Library, 6th ser., 13
[1991], 297–322 [p.
318].) When there is a watermark, its position in
the sequence should be shown,
using Stevenson's system (explained in
my essay); and the
sequence should reflect the sheet as viewed from the
mould side with the watermark
right side up (the standard orientation for
bibliographical use: see below for Needham's use of it). Obviously one
to work with the side showing the obverse of the image. When a water-
mark is not present, the orientation is not significant, since the sequence
serves its purpose whether read forward or backward. The great value of
giving all the chain-space measurements is that all handmade papers can
thereby be identified, even when they have no watermarks or when only
an unwatermarked part-sheet is present (as in some half-sheet gather-
ings). In the first of the two sample descriptions near the end of my es-
say, the identification of the watermark and countermark should include
only the height and width measurements and should be followed (after a
semicolon) by the chain-space sequence, incorporating a notation of the
water mark position.
The most extensive use, thus far, of Vander Meulen's
observations
has been made by David L. Gants and R. Carter Hailey. Although their
methods of collecting data
differ (Gants's is high-tech and Hailey's low-
tech), they both make chain-space
sequences the center of their paper
descriptions—accompanied by indications of
wireline density (see below)
and measurements of watermarks, with digital
photographs (Gants) or
freehand drawings (Hailey) of the watermarks. Both follow
Vander Meu-
len's plan of using vertical lines to
separate chain-space measurements,
but each makes some modifications. Gants, when
explaining his approach
to building a database, says that he records the sequence
for every sheet
(not a representative one for any sheet from the same mould) and
that he
takes his measurements from either side of the watermark (not the
center
of the sheet), as a location accessible in a variety of formats. See
"Identi-
fying and Tracking Paper Stocks in Early Modern London," in Papers of
the Bibliographical Society of
America
, 94 (2000), 531–540. However, in
his
online "A Digital Catalogue of Watermarks and Type Ornaments Used
by William Stansby [1614–1618]" (2005), he
does not report the sequence
for each sheet individually, and he gives two
chain-space sequences per
mould, from the "top" and "bottom" of the sheet (that is,
near the two
longer edges, when the watermark is viewed in upright position). He
also
uses different symbols from Vander Meulen's:
parentheses for chainlines
that intersect watermarks, braces for deckle edges,
double slashes for gaps
in measurement, and square brackets for trimmed edges.
Hailey makes
fewer alterations to Vander Meulen's form of
report, the chief one being
to print in bold-face italics the figures for the
space(s) where a watermark
is located; and each of his chain-space measurements is
an average of the
measurements from several sheets, the result being a "composite
chain-
space model." (In using this approach, one must be careful not to
obscure
differences between paper stocks.) His method is set forth in detail
on
pp. 156–165 of "The Shakespearian Pavier Quartos Revisited" (Stud-
ies in Bibliography, 57 [for 2005–06; 2008],
151–195); he had previously
tique Laid Paper: A Method" (in Teaching Bibliography, Textual Criticism,
and Book History, ed. Ann R. Hawkins [2006], pp. 149–154) and in "The
Dating Game: New Evidence for the Dates of Q4 Romeo and Juliet and
Q4 Hamlet" ( Shakespeare Quarterly, 58 [2007], 367–387 [see pp. 372–373,
376–377]).
Vander Meulen's
1984 essay also discusses several
other ways to iden-
tify paper, none of which is covered in my essay. Their
inclusion in a
bibliographical description, however, is not necessarily required.
The one
most likely to qualify is a measurement of wireline density—expressed
(as
Vander Meulen suggests) in the number of wires per
three centimeters.
(A recent use of the three-centimeter measurement, reported to me
by
Vander Meulen, occurs in Agnieszka
Helman-Ważny's
The Archaeology of
Tibetan Books [2014], as on
pp. 35, 177, and 231–248.) Since paper can
be identified by the chainline sequences
alone, the addition of wireline
density may not be of further help, though it can
sometimes be useful
in making tentative discriminations among paper varieties and in
deter-
mining which varieties come from the same pair of moulds. In any case,
a
thorough description is of course not limited to details necessary
for
identification, and the inclusion of wireline density obviously adds to
the
completeness of the description. Another identifying characteristic is
the
precise location of any tranchefiles (a tranchefile is an extra chain
that
sometimes occurs, at one or both ends of the mould, between the
last
regular chain and the frame of the mould); but since these lines would
be
part of the records of chain-space sequences (and recognized by the
smaller
intervals they usually create), they do not normally need separate
attention.
Still another feature of laid paper is the "shadow" centered on many
chainlines,
caused by the supporting wooden ribs under the chains (ex-
cept for most
tranchefiles). Sometimes, apparently when chains are not
directly aligned over the
ribs (as they normally are), there is an effect of
double chainlines, one of which
in each pair is less distinct than the other.
Vander
Meulen does not include these secondary chainlines in his chain-
space
sequences, nor does he report the location of shadows in relation to
the secondary
chainlines. Although the presence of these features should
perhaps be mentioned in
general terms in a bibliographical description,
they do not usually call for
detailed reporting. A further article of Vander
Meulen's,
published four years after this major one, should be mentioned
here as a footnote to
it: "The Low-Tech Analysis of Early Paper" (Literary
Research, 13 [1988], 89–94), which conveniently sums up
the techniques
for analyzing paper and the literary and historical uses to which
such
analysis can be put.
Paul Needham has published an important series of essays on
the
history of paper and the analytical uses of paper evidence. Of these,
the
two that are perhaps most directly relevant to bibliographical descrip-
tion
are "The Paper Supply of the Gutenberg Bible" (Papers of the
Biblio-
graphical Society of America
, 79
[1985], 303–374) and "Allan H.
Stevenson
and the Bibliographical Uses of Paper" (Studies
in Bibliography, 47 [1994],
23—64). The former includes the
following example of a form for record-
ing a paper stock (p. 317), which is
repeated with further explanation and
commentary in the latter (pp. 32—33, and with
the addition of the "a"
and "b"):
Royal: Bull's Head
a. mR4 | 19 | 14 chains (ave. 42.1 mm) | 18 | (6+)
b. mL4 (3+) | 21 | 14 chains (ave. 42.3 mm) | 18 | (1+)
The heading shows the common name for the size of the sheet (Royal),
along with a
descriptive term for the type of watermark (Bull's Head),
without further
specification of the dimensions of either one. Needham
feels that precise leaf measurements of handmade paper are not
useful
even for an example with surviving deckle edges, given the variations
that
would exist between measurements taken at different points and the
dif-
ficulty of estimating how much is inaccessible in the fold. As long as
the
approximate nature of such measurements is clear, however, there is no
harm
in providing them as evidence for the extrapolation to the named
size. The "a" and
"b" lines in the description refer to the twin moulds of
the particular paper stock;
and "mR" and "mL" give the location of the
watermark as being in the right or left
half of each mould, with the at-
tached number indicating which chain-space it is in
(here, the fourth from
the nearest short edge). For "right" and "left" to be
meaningful, of course,
the paper must be consistently viewed from the same side and
with the
watermark image upright. Needham recommends
looking at the mould
side; and if in certain instances the felt side seems more
appropriate, the
designation "(f.s.)" can be added. When a watermark design has no
obvi-
ous top or bottom, the position of some arbitrary feature can sometimes
be
used and noted (though totally symmetrical abstract designs defeat the
system
altogether).
The remainder of each "a" and "b" line gives, at each end in paren-
theses, the
distance from the deckle edge to the tranchefile; then, moving
in from each end, the
number between the vertical rules is the measure-
ment from the tranchefile to the
nearest regular chain; and finally, in
the center, is the number of chainlines with
the average distance be-
tween them. (One possible ambiguity—whether or not the
stated num-ber of chains includes the two apparently represented by the second and
spaces" and not to "chains.") Needham acknowledges that this statement
of chainline numbers is "a very crude 'measurement,'" but he believes it
is usually sufficient for paper-stock identification when used in conjunc-
tion with the other details. He recognizes, however, that noting the size
of every chainline interval, as suggested by Vander Meulen (see above),
may sometimes be needed, especially in connection with unwatermarked
paper. The tabular form of Needham's presentation is obviously not es-
sential: his notation can be made to fit within the general scheme I sug-
gested at the end of my essay.
One notices that Needham's description does not include an
indi-
cation (using Stevenson's system or any other) of
the exact position of
a watermark in relation to the adjacent chainlines. The reason
is that
Needham feels the information conveyed by such
measurements is
"so inferior to that supplied by actual-size reproductions" that he
can-
not regard them as "a fundamental element of paperstock description"
(p.
34)—though he notes that they can indeed be helpful, as Vander
Meulen has shown. In any case, reproductions cannot always be
sup-
plied; and, even when they are, some bibliographers will feel that
they
should be supplemented by measurements. But it is true, of course,
that
reproductions will always provide details not covered by verbal
accounts.
Needham's essay on Stevenson includes other criticisms, but they are
subordinate to the
admiration expressed for Stevenson's overall
achieve-
ment. Reading Needham's assessment of Stevenson should be an impor-
tant part of any bibliographer's
preparation for thinking about paper.
Several other essays of his also make basic
contributions to this process:
"Res papirea: Sizes and Formats of the Late Medieval
Book," in Ration-
alisierung der Buchherstellung im Mittelalter
und in der frühen Neuzeit, ed. Pe-
ter Rück and
Martin Boghardt (1994), pp. 123–145; "Concepts
of Paper
Study," in Puzzles in Paper: Concepts in Historical
Watermarks, ed. Daniel
W. Mosser, Michael Saffle, and Ernest W. Sullivan
II (2000), pp. 1–36;
and his superb survey "The Paper of
English Incunabula," in Catalogue of
Books Printed in the XVth
Century Now in the British Library, ed. Lotte
Hel-
linga (2007), pp. 311–334.
B. J. McMullin's articles of 2003–04, 2004, and
2008 (cited above,
under "Format") are basic for the study of
handmade wove paper in
the late eighteenth and early nineteenth centuries and
machine-made
paper in the first half of the nineteenth century. They provide
historical
background and explain some analytical techniques applicable to
those
papers. In addition to outlining clues regarding format (see
"Format"
above), he shows (in the 2008 paper) how seam marks in
machine-made
paper can be used for dating (paper with seam marks is probably
between
1810 and 1850) and for identifying cancels and
other irregularities in the
tern). An unanswered historical question that hampers some analysis is
whether the printer gathered and collated sheets before sending them to
the binder or whether, in the era of publishers' bindings, the binder took
on this job; if the latter, random combinations of part-sheets in finished
books might be more likely, thus rendering more doubtful the determina-
tion of the particular imposition used. The point most relevant to writing
a bibliographical description is his suggestion of two alternative plans for
recording seam marks (pp. 78–79):
seam marks horizontal: B1 2 7 8 F3 4 5 6 H3 4 5 6 M1 2 7 8
seam marks vertical: C5 8 G6 7 K6 7 N5 8 R1 4 U2 3 Z2 3
seam marks horizontal: $1 2 7 8 BM, $3 4 5 6 FH
seam marks vertical: $1 4 R, $2 3 UZ, $5 8 CN, $6 7 GK
As he points out, however, the pattern will not necessarily be the same in
every
copy of the edition; and if one therefore inserts library sigla (or copy
numbers) in
parentheses to identify individual copies, he feels that the first
system would
probably be clearer—though I believe that the preferable
one might vary in different
situations, since each has a different focus.
In any case, they are both usable;
and, though other schemes could be
devised (especially to give more information
about the exact positioning
of the seam marks), there would seem to be no reason in
most instances
to do so.
I do, however, think there is reason to question the point McMullin
makes next: that "one might well allow the record to stand
unqualified
[with copy sigla], to be understood as indicating that a seam mark of
a
particular orientation is to be found in the specified gathering in at
least
one copy from within the edition." One could, of course, adopt an
ab-
breviated form of paper description and omit the locations of seam
marks
entirely (saying simply "machine-made paper with seam marks"); but if
one
is going to specify a sample location, the copy (or copies) used to
obtain that
information should be specified. The situation is analogous to
the recording of
locations of other features that vary from copy to copy.
In the case of handmade
paper, for example, all the sheets of a book may
have the same watermark and
countermark, but those marks may not
appear on the same leaves in every copy,
depending on how the paper
was turned before printing and how the printing and
handling of half-
sheet gatherings were carried out; furthermore, the small
variations in the
papers coming from twin moulds (variation in watermark placement
and
in chainline and tranchefile spacing) may not be identically distributed
in
every copy. In all such cases, therefore, one ought to cite not only
signa-
ture locations but also the specific copies used—and normally to do so in
copies examined (see "Arrangement" below), since what is being recorded
is a feature of the edition as published, not a post-publication alteration.
A justifiable exception to this last recommendation, however, can be illus-
trated by Stanley Boorman's bibliography of Ottaviano Petrucci (2006),
where he (remarkably) records the locations of watermarks throughout
each copy examined (that is, not simply sample locations); this information
(along with his thorough notes on other variable features, such as in-house
corrections) is much more efficiently placed in the accounts of examined
copies, which thus form the largest element in his descriptions.
One other matter that McMullin has written about is the
paper-quality
mark—the symbol that may appear in the direction line on the first
page
of a gathering to identify the paper when two issues on different
papers
are produced. The presence of such marks should obviously be noted in
the
paragraph on paper in a description. An example of these marks is
illustrated by
McMullin in What Readers Should Ignore on
the Printed Page
(2014), p. 7; see also his "Paper-Quality Marks
and the Oxford Bible
Press, 1682–1717," The
Library, 6th ser., 6 (1984), 39–49, and "Cowper's
Complete Poetical Works, 1837
(Russell, 166)," Script & Print, 40
(2016),
45–54 (esp. pp. 51–53).
John Bidwell's contributions deal primarily with the history
of pa-
permaking and thus provide useful background for descriptive
bibliogra-
phers, even if they are not always relevant to the practice of
description.
(For an excellent overview, see especially his "The Study of Paper
as
Evidence, Artefact, and Commodity," in The Book
Encompassed, ed. Peter
Davison [1992],
pp. 69–82.) But descriptive bibliographers will be glad
to find in his extraordinary
American Paper Mills, 1690–1832: A Directory of
the
Paper Trade, with Notes on Products, Watermarks, Distribution Methods,
and
Manufacturing Techniques (2013) a list of watermark attributions, citing
each
watermark and naming the mill that used it (pp. lxx–lxxiii). They will
also
be happy to find in "The Size of the Sheet in America: Paper Moulds
Manufactured by N. & D. Sellers of Philadelphia"
(Proceedings of the Ameri-
can Antiquarian Society, 87
[1977], 299–342) a table headed "Some Names
for Anglo-American Paper
Sizes and Their Measurements." What makes
this table more useful than others of its
kind is that, for each of fifteen
named sizes, the dimensions reported at nine
specified times (from 1713
to 1952) are given.
Bidwell's principal exercise in actual description of paper
occurs in
his Fine Papers at the Oxford University Press
(1999), where (after an excel-
lent historical account of the
manufacture and sale of handmade paper)
he provides fascinating commentaries on the
forty specimens that are
included in the book. Besides identifying these papers, he
gives a concise
would have recorded in its ledgers (but recognizing that such descriptions
did not have to be self-sufficient because papermakers would have been
shown samples of what was desired). Each of his descriptions consists of
seven elements: weight (per ream of 480 sheets), texture (laid or wove),
dimensions, substance (grams per square meter), thickness (caliper mea-
surements in thousandths of an inch), color, and watermark. Here is an
example: "28–pound hand-made antique laid, 16¼ × 22½ ins: 106 gsm,
caliper .005 in., cream, watermarked near each of the four corners: W [flower]
M (felt side)" (p. 46). Clearly two of these features, weight and substance,
could not be reported from direct observation in descriptions of books
since they require loose sheets (though they might be provisionally noted
from external sources); but the other five are part of the standard I sug-
gested in 1971, though presented differently. Bidwell recognizes that the
thickness of a sheet of handmade paper is likely to vary from one part
of the sheet to another, and thus the significance of the thickness figures
he gives is not clear (a point made by David L. Vander Meulen in his
review of this book in Papers of the Bibliographical Society of America , 97
[2003], 589–595). In my 1971 essay I suggested that such figures should
be expressed either as a range or as an average; I would now add that,
in the latter case, the word "average" or the abbreviation "avg." should
be appended to emphasize the status of the figure. (An article that may
be regarded as a supplement to Bidwell, though published earlier, is John
Purcell's "The Availability of Hand-Made, Mould-Made and Fine Ma-
chine-Made Papers," Matrix, 3 [1983], 67–75, which includes a three-
page table giving specifications for twenty-nine papers that were available
at that time.)
One of Bidwell's services to descriptive bibliographers is to
have writ-
ten a thorough review of Peter F. Tschudin's
Grundzüge der Papiergeschichte
(2002) for Papers of the Bibliographical Society of America
(98 [2004], 105–
109). Because Tschudin recommends describing paper and watermarks
in terms
of the International Standard for the Registration of Paper with
or
without Watermarks (first proposed by the International Association of
Pa-
per Historians [IPH] in 1992 and last revised in 2013
[text available on
the IPH website]), Bidwell had occasion
to assess that system—a task
that most bibliographers have not undertaken. The
system is designed to
produce machine-readable records, each of which has
potentially a hun-
dred fields, sometimes involving intricate coding. As an example,
Bidwell
notes that a watermark consisting of a crown
over CLK in script would
be reported as "R3/1—{b:(i:X"CLK")}"—and this is far from
being one
of the more complex possibilities. Although a large database
constructed
in this way could indeed be searched productively, the effort involved
in
comes to the sensible conclusion that descriptive bibliographers may justi-
fiably feel that their purposes are adequately served by readily understood
verbal watermark descriptions (including, as Stevenson suggested, quasi-
facsimile transcriptions).
A more detailed criticism of the IPH system (referred to by Bidwell)
has been provided on the website produced at Bates College by
Robert W.
Allison and James A.
Hart, "The WWW Watermark Archive Initiative,"
which is aimed at developing
guidelines for online databases: see the sec-
tion entitled "Commentary and
Interpretation of the IPH Standard." In
their comments on watermarks, they say, "Our
objective is to use straight-
forward standardized language and eliminate the
codes." This website
also contains links to several online databases of watermark
images, such
as the University of Delaware's "Thomas
Gravell Watermark Collection,"
containing some 8500 records as of early
2015; and the IPH website lists
more than a dozen such links. (See
below for some further references to
discussions of online databases.)
The next year after my essay, Philip Gaskell (whose important
earlier
studies of paper are mentioned there) published A New
Introduction to Bib-
liography (1972, 1974), which
discusses the history and analysis of paper in
two chapters. The one on paper in the
hand-press period (pp. 57–77) has
a page on description (pp. 76–77), and that on the
machine-press period
(pp. 214–230) has two pages (pp. 226–228). The former chapter
includes
the point, which cannot be stated too often, that handmade paper
varies
from sheet to sheet and that any edition may have been printed on two
or
more paper stocks, which may not be consistently represented in a given
copy
(emphasizing once again the necessity for examining a large number
of copies). But
the main contribution to paper description that Gaskell
makes in this book has to do with machine-made paper. He
enumerates
five simple tests for distinguishing the characteristics of different
stocks:
(1) assess the feel (surface texture) and color in a good light; (2)
identify,
with a raking light, the belt side and notice the wove pattern; (3)
measure
the pattern (in wires per centimeter) and, if it is oblong or
diamond-shaped
(rather than square), notice whether its longer dimension is parallel
to
the grain (which is also the machine direction); (4) measure the
thickness
with a micrometer; and (5) establish the relative densities of papers in
two
books by weighing them (though it should be added that there are often
too
many variables here to make this test of much use). He then lists seven
additional
tests that, because they involve damaging the paper (by fold-
ing or applying
chemicals), would not normally be of use to descriptive
bibliographers. Tables of
standard paper sizes are given for the hand-press
period (pp. 73–75; provided in my
essay as well) and the machine-press
Fleeman's bibliography of Samuel Johnson [2000], p. xxxvii.)
Besides Gaskell, another summation of basic information about
pa-
per (far more extensive than Gaskell's), incorporating
original research
and with fresh observations, is Neil
Harris's
Paper and Watermarks as Bib-
liographical Evidence
(2010, 2017, on the website of the Institut
d'Histoire
du Livre in Lyon; a pdf of the second edition
can be printed out as a
document of 155 very full pages). It is a learned treatise
on many aspects
of paper study, presented in a "snappy" and "punchy" way (his
words),
with frequent jocular asides. (This style is not to everyone's taste,
but
the work does contain some memorable sentences.) The most relevant
parts for
descriptive bibliographers are embedded in the fourth and fifth
chapters: "The Shape
of Paper" (pdf pp. 32–44), which deals with size
and format; and "Dillying and
Dallying with Watermarks" (pp. 45–59),
which includes comments on describing,
reproducing, and classifying
watermarks. Bibliographers who have occasion to consult
published col-
lections of watermark designs will benefit from the fascinating
chapter on
Briquet (pp. 60–75), which takes up his life, methods, and followers,
as
well as how to employ his great work. In using any of Harris's discussions,
one would be well advised to read, at the same time,
the corresponding
sections of his impressive seventy-page analytical record of the
interna-
tional literature, full of references not easily found elsewhere and
advice
not available anywhere else—see especially sections 7–23 (pp. 112–131,
on
size, format, and watermark description and reproduction), 30–31
(pp. 136–144, on
analytical bibliography and watermark collections), and
35 (pp. 150–152, on
websites).
What may be regarded as a supplement to Gaskell's and McMul-
lin's work on nineteenth-century paper is Chris Elmore's "Describing
Nineteenth-Century Papers" (Script & Print, 40 [2016], 5–28), though
its
emphasis is on writing papers rather than printing papers (it should
certainly be
consulted by those dealing with handwritten letters and
journals). Of the points
relevant to bibliographers of printed books, two
are worth repeating here, even
though they are obvious to anyone who
understands that paper made with the
Fourdrinier machine has a wove
pattern, imparted by the moving woven-wire belt on
which the pulp is
placed, and that any chainlines or watermarks result from the
action of a
dandy roll pressing into the other side of the pulp. Knowing this
process,
one can tell whether paper with a laid pattern is machine-made by
noting
whether it has a wove pattern as well as a laid pattern—the former
of
course indented into
one side of the paper, and the latter indented
into
the other. (The indentations from the wove belt were deeper and
closer
together than those resulting from hand-held moulds, but this point,
watermarked paper (even without laid lines) can naturally be detected
by the presence of the watermark indentations on the opposite side from
the wove-pattern indentations; but one can also note that the indenta-
tions of the lines of the watermark design (and lettering, if any) are inter-
sected by the indentations of the wove pattern from the belt—whereas
in handmade wove paper the watermark, being on top of the wire mesh,
covers the mesh pattern at the points of contact. (One must recognize
that these facts about paper made with the Fourdrinier machine do not
apply to so-called mechanical mould-made paper, in which a rotating
cylinder mould deposits each sheet on a moving woolen felt, and in which
all patterns and watermarks are therefore indented from the same side;
another difference between such paper and Fourdrinier paper is that its
fibers are distributed randomly in the sheet rather than solely in the belt
direction.) Employing contemporary paper-trade sources, Elmore says
that "laid" and "wove" were the standard nineteenth-century terms for
printing papers. As for dimensions, he gives tables of writing-paper sizes
(and size names), but he does not add anything to Gaskell's table of ma-
chine-made paper sizes. Although he suggests that bibliographers might
use some of the trade terms for other characteristics (such as finish, bulk,
and opacity), he also indicates that they were somewhat impressionistic,
reflecting the judgment of professionals who handled paper every day,
and that they should therefore be used with caution—or, I would add,
probably not be used at all (or only as ordinary adjectives, not as techni-
cal terms).
Among the many other post-1971 publications on the history
and
analysis of paper, several may be mentioned as particularly useful
for
descriptive bibliographers (and see the references in "Format"
above).
Published collections of watermark reproductions have improved
since
1971 as a result of the use of beta-radiography and several
photographic
and digital processes. (See David E.
Schoonover, "Techniques of Repro-
ducing Watermarks: A Practical
Introduction," in Essays in Paper Analysis,
ed. Stephen Spector [1987], pp. 154–167; A. de la Chapelle, C. Monbeig-
Goguelle, and A. Prat, "Les filigranes des dessins anciens et les
relèves
betaradiographiques," Annals of Radiology, 37
[1994], 249–258; David L.
Gants, "The
Application of Digital Image Processing to the Analysis of
Watermarked Paper and
Printers' Ornament Usage in Early Printed
Books," in New Ways of
Looking at Old Texts, II, ed. W. Speed Hill [1998],
pp. 133–147;
and Neil Harris, Paper and Watermarks as
Bibliographical Evi-
dence [2017], pp. 54–57, 128–131.) For
example, there are Thomas L.
Gravell and George Miller's two volumes (A Catalogue of
American Water-
marks, 1690–1835 [1979; rev. with
Elizabeth Walsh, 2002] and A Catalogue
of Foreign
Watermarks Found on Paper Used in America,
1700–1835
[1983]);
Thomas L. Gravell Watermark Archive on the Internet," in Puzzles in
Paper: Concepts in Historical Watermarks, ed. Mosser, Sullivan, and Michael
Saffle (2000), pp. 211–228. A notable publication by a great cartographic
scholar, David Woodward, is his Catalogue of Watermarks in Italian Printed
Maps ca. 1540–1600 (1996). A number of collections of watermarks are
now available online, and two websites that conveniently provide links to
many of them are those of the International Association of Paper Histo-
rians and of the "Bernstein: The Memory of Paper" consortium; see also
Harris (cited three sentences earlier), pp. 150–152.
The two anthologies just cited, Spector (1987) and Mosser
(2000), in-
clude a number of other pieces relevant to the
description and recording
of watermarks. The Spector contains Phillip Pulsiano's
extremely use-
ful list, "A Checklist of Books and Articles Containing
Reproductions
of Watermarks" (pp. 115–153). The Mosser presents three articles
on
watermark reproduction: Carol Ann Small's
"Phosphorescence Water-
mark Imaging" (pp. 169–181), Rolf
Dessauer's "DYLUX, Thomas L.
Gravell, and
Watermarks of Stamps and Papers" (pp. 183–185), and Dan-
iela
Moschini's "La Marca d'Acqua: A System for the Digital Recording
of
Watermarks" (trans. Conor Fahy; pp. 187–192). Also in the
Mosser
is Robert W. Allison's "An Automated World Wide Web
Search Tool
for Papers and Watermarks: The Archive of Papers and Watermarks
in
Greek Manuscripts" (pp. 201–210), which describes the Bates College
archive
mentioned above; and Ted-Larry Pebworth's "Towards a
Tax-
onomy of Watermarks" (pp. 239–242), which offers a computer-oriented
system
based on a "flexible grid pattern" for describing locations of parts
of
watermarks.
A third anthology, Looking at Paper: Evidence &
Interpretation, ed. John
Slavin et al.
(2001), emphasizes paper in prints, drawings, and manuscripts
and has
less of relevance for descriptive bibliography; but it does include
Ruby Reid Thompson's "Historical and Literary Papers and the
Applica-
tion of Watermark Descriptions" (pp. 142–153), which makes use of
the
Nottingham University Library Watermark Database, and Ian
Christie-
Miller's "Digital Imaging" (pp. 139–141), on the "Bookmark"
reflected-
light system. A later article on Christie-Miller's approach is his "New
Tools for Old Paper" (The Book Collector, 58 [2009], 383–389),
which
describes his Advanced Paper Imaging System, using both frontlighting
and
backlighting and showing conjugate leaves together. The equipment
that Christie-Miller has devised is illustrated on his website
(www.early-paper.com); the battery-powered one-millimeter-thick
electroluminescent
source for viewing paper structure and watermarks, which is a
part of
his system, has been marketed as "Pocket Viewlight" by Howard
Eaton
Lighting Ltd.
The series of articles reporting on the use of the Davis
(California)
cyclotron to analyze the makeup of paper and ink is most
conveniently
represented by Richard N. Schwab, Thomas A. Cahill,
Bruce H. Kusko,
and Daniel L. Wick's "Cyclotron
Analysis of the Ink [and paper] in
the 42–Line Bible," Papers of
the Bibliographical Society of America
,
77
(1983), 285–315. Two good general articles on the usefulness of
water-
mark evidence, both published in 1978, are Bruno Scarfe's "A Role for
Watermarks in Bibliographical Description, with
Special Reference to a
Collection of Spanish Dramatic Items," Bibliographical Society of Australia
and New Zealand Bulletin, 12: 85–101; and Stephen Spector's "Symmetry
in Watermark Sequences," Studies in Bibliography, 31: 162–178. Further
references can be found in my
Introduction to Bibliography (2002
revision),
part 5, pp. 181–193, and sections 9D5, 9F2, 9G5, and 9H4 (pp.
205–207,
290–291, 307–309, and 314).
To summarize: the main points to be added to my 1971 recommenda-
tions
are that one would be well advised to record the sequence of chain-
line intervals
in a stock of paper; that any indication of the location of a
watermark should
specify which half of the sheet it is in; that any citation
of a given leaf as a
source of evidence should indicate the particular copy
of the book used; and that
specifications of thickness of handmade paper
should note that the figure is an
average. These additions will make the
treatment of paper both more comprehensive
and more precise.
| Notes on Recent Work in
Descriptive Bibliography | |